Skip to content
Model Pulse

Hourly diffs

Changelog

Every detected change to the model landscape: releases, price moves, context windows, capabilities and deprecations. Also available as RSS or JSON.

Change type

Time range

4038events

22 events
repriced
DeepSeek V4.1 Flash

via openrouter

What changed

  • Input price$0.15/M$0.3/M▲ 100.0%
  • Output price$0.6/M$1.20/M▲ 100.0%
  • Cache read$0.0030/M$0.0060/M▲ 100.0%
repriced
DeepSeek V4 Pro 0813 (Alibaba)

via edenai

What changed

  • Input price$1.12/M$0.581/M▼ 48.2%
  • Output price$3.37/M$1.74/M▼ 48.2%
context
GLM-5.2

via kilo

What changed

  • Context203K1.05M5.2×
  • Output limit182K131K−28%
repriced
Kimi Latest

via openrouter

What changed

  • Cache read—$0.23/M
repriced
Schematron V2 Small

via openrouter

What changed

  • Cache read—$0.05/M
repriced
MoonshotAI: Kimi Latest

via kilo

What changed

  • Cache read—$0.23/M
context
Nemotron 3 Ultra 550B A55B

via kilo

What changed

  • Context256K203K−21%
  • Output limit33K183K5.6×
repriced
GLM-5.3

via openrouter

What changed

  • Output limit131K944K7.2×
  • Input price$1.09/M$1.40/M▲ 28.2%
  • Output price$3.43/M$4.40/M▲ 28.2%
  • Cache read$0.203/M$0.26/M▲ 28.2%
context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.05M1.02M−2%
  • Output limit393K384K−2%
repriced
Schematron V2 Turbo

via openrouter

What changed

  • Cache read—$0.03/M
repriced
DeepSeek V4 Flash 0731 (Alibaba)

via edenai

What changed

  • Input price$0.352/M$0.176/M▼ 50.0%
  • Output price$1.06/M$0.528/M▼ 50.0%
repriced
Qwen3 14B

via openrouter

What changed

  • Output limit8K16K2×
  • Input price$0.228/M$0.12/M▼ 47.3%
  • Output price$0.91/M$0.24/M▼ 73.6%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.04/M$0.06/M▲ 50.0%
  • Output price$0.08/M$0.12/M▲ 50.0%
  • Cache read$0.0080/M$0.012/M▲ 50.0%
context
Qwen: Qwen3 14B

via kilo

What changed

  • Context131K41K−69%
  • Output limit8K16K2×
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Output limit393K384K−2%
  • Input price$0.578/M$1.05/M▲ 82.4%
  • Output price$1.73/M$3.16/M▲ 82.4%
  • Cache read$0.018/M$0.035/M▲ 91.1%
repriced
Inference.net: Schematron V2 Small

via kilo

What changed

  • Cache read—$0.05/M
repriced
Nemotron 3 Ultra 550B A55B

via openrouter

What changed

  • Output limit33K183K5.6×
  • Input price$0.625/M$0.6/M▼ 4.0%
  • Output price$3.13/M$2.40/M▼ 23.2%
  • Cache read$0.188/M$0.12/M▼ 36.0%
repriced
Inference.net: Schematron V2 Turbo

via kilo

What changed

  • Cache read—$0.03/M
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.049/M$0.09/M▲ 83.4%
  • Output price$0.098/M$0.18/M▲ 83.4%
  • Cache read$0.0098/M$0.018/M▲ 83.4%
new model
DeepSeek V4.1 Flash (Consensus Protocol)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
GLM-5.2

via openrouter

What changed

  • Output limit182K131K−28%
  • Input price$0.6/M$0.683/M▲ 13.9%
  • Output price$2/M$2.15/M▲ 7.4%
  • Cache read$0.15/M$0.127/M▼ 15.4%
context
GLM-5.3

via kilo

What changed

  • Context1.05M1.05M−0%
  • Output limit131K944K7.2×

329 events
capability
Nova 2 Lite (JP)

via amazon-bedrock

What changed

  • last updated2025-12-022025-12-01
capability
Gemma 3 12B IT

via openrouter

What changed

  • nameGemma 3 12BGemma 3 12B IT
  • knowledge2024-08-312024-08
  • release date2025-03-132025-03-12
  • last updated2025-03-132025-03-12
repriced
DeepSeek V4.1 Flash (Deep Infra)

via edenai

What changed

  • Input price$0.3/M$0.2/M▼ 33.3%
  • Output price$1.20/M$0.6/M▼ 50.0%
new model
Gemma 3 4B

via merge-gateway

What changed

First observed in the model catalog

repriced
GLM-4.7

via ofox

What changed

  • Output price$2/M$2.20/M▲ 10.0%
  • Cache read$0.08/M$0.11/M▲ 37.5%
repriced
DeepSeek V4 Flash 0423

via ofox

What changed

  • Input price$0.44/M$0.19/M▼ 56.8%
  • Output price$1.32/M$0.51/M▼ 61.4%
  • Cache read$0.014/M$0.028/M▲ 100.0%
capability
Qwen3-Coder 30B-A3B Instruct

via amazon-bedrock

What changed

  • nameQwen3 Coder 30B A3B InstructQwen3-Coder 30B-A3B Instruct
  • descriptionQwen coding model for software agents, repository edits, and code reasoningSmaller Qwen coder for efficient local agents and repo-level fixes
  • open weightsNoYesEnabled
  • knowledge2024-042025-04
  • +1 more changes
new model
Kimi K3 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Gemma 3 27B IT (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

capability
Llama 3.1 8b Instruct

via nano-gpt

What changed

  • tool callNoYesEnabled
capability
Llama-3.3-70B-Instruct

via tinfoil

What changed

  • attachmentYesNoRemoved
capability
Mistral Large 3

via merge-gateway

What changed

  • release date2024-11-012025-12-02
capability
Ministral 14B 3.0

via amazon-bedrock

What changed

  • descriptionCompact Mistral model for edge, latency-sensitive, and cost-efficient workloadsOpen vision-language model for efficient local deployment, instruction following, and tool use
  • attachmentNoYesEnabled
  • open weightsNoYesEnabled
  • release date2024-12-012025-12-02
  • +2 more changes
new model
GLM-5V-Turbo

via llmgateway

What changed

First observed in the model catalog

context
Llama 4 Scout 17B Instruct

via amazon-bedrock

What changed

  • descriptionOpen multimodal Llama model for long-context analysis and efficient agentsOpen Llama with long-context vision for efficient multimodal agents
  • Context3.50M10M2.9×
  • Output limit16K8K−50%
new model
GLM-5.3

via agentrouter

What changed

First observed in the model catalog

new model
Voxtral Mini 3B 2507 (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

new model
Gemma 3 4B IT

via deepinfra

What changed

First observed in the model catalog

repriced
GPT Pro Latest (GPT-5.5 Pro)

via edenai

What changed

  • Cache read$3/M—
  • tiers[object Object][object Object]
new model
inclusionAI: Ling 3.0 Flash VL

via kilo

What changed

First observed in the model catalog

repriced
Qwen3 VL 235B A22B Instruct

via amazon-bedrock

What changed

  • nameQwen/Qwen3-VL-235B-A22B-InstructQwen3 VL 235B A22B Instruct
  • descriptionQwen vision-language model for visual reasoning, documents, and agent tasksQwen vision-language instruct model for visual reasoning, documents, and agent tasks
  • open weightsNoYesEnabled
  • knowledge—2025-03-31
  • +4 more changes
capability
GPT OSS Safeguard 20B

via amazon-bedrock

What changed

  • reasoningNoYesEnabled
context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
new model
DeepSeek V4 Flash (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
gpt-oss-120b

via amazon-bedrock

What changed

  • descriptionOpen-weight GPT model for self-hosted reasoning and instruction-following workloadsOpen GPT reasoning model for self-hosted agents and controllable deployments
context
Llama 4 Scout 17B Instruct (US)

via amazon-bedrock

What changed

  • Context3.50M10M2.9×
  • Output limit16K8K−50%
new model
Hy-MT2 Plus

via llmgateway

What changed

First observed in the model catalog

capability
Llama-3.3-70B-Instruct

via abacus

What changed

  • attachmentYesNoRemoved
context
MiniMax-M2.5

via kilo

What changed

  • Context205K200K−2%
  • Output limit131K128K−2%
new model
DeepSeek V4.1 Flash (Together AI)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Ministral 3 8B

via amazon-bedrock

What changed

  • descriptionCompact Mistral model for edge, latency-sensitive, and cost-efficient workloadsCompact open vision-language model for edge deployment, instruction following, and tool use
  • attachmentNoYesEnabled
  • open weightsNoYesEnabled
  • release date2024-12-012025-12-02
  • +2 more changes
repriced
Qwen3 Coder Next

via amazon-bedrock

What changed

  • descriptionQwen coding model for software agents, repository edits, and code reasoningOpen-weight Qwen coding model for agents, repository edits, and multi-turn tool use
  • reasoningYesNoRemoved
  • knowledge—2025-09
  • release date2026-02-062026-02-03
  • +3 more changes
context
Qwen3.5 35B A3B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
new model
Gemma 3 27B

via empiriolabs

What changed

First observed in the model catalog

capability
Mistral Large 3

via amazon-bedrock

What changed

  • descriptionFlagship Mistral model for advanced reasoning, coding, and multilingual workMistral's largest general model for enterprise agents, coding, and multilingual reasoning
  • familymistralmistral-large
  • attachmentNoYesEnabled
  • knowledge—2024-11
context
Mistral Small 3.2 24B

via openrouter

What changed

  • Context131K256K2×
capability
Mistral Large 3

via anyapi

What changed

  • release date2024-11-012025-12-02
capability
Gemma 3 12B IT

via kilo

What changed

  • nameGoogle: Gemma 3 12BGemma 3 12B IT
  • open weightsNoYesEnabled
  • knowledge—2024-08
  • release date2025-03-132025-03-12
  • +1 more changes
new model
Gemma 3 27B IT (Nebius)

via edenai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731 (Alibaba)

via edenai

What changed

  • Input price$0.176/M$0.352/M▲ 100.0%
  • Output price$0.528/M$1.06/M▲ 100.0%
context
Palmyra X5 (US)

via amazon-bedrock

What changed

  • Input limit—1.04M
repriced
GLM Latest

via openrouter

What changed

  • Output limit944K236K−75%
  • Input price$1/M$0.936/M▼ 6.4%
  • Output price$3.41/M$3.17/M▼ 7.1%
  • Cache read$0.2/M$0.187/M▼ 6.4%
capability
Gemma 3 27B IT

via kilo

What changed

  • nameGoogle: Gemma 3 27BGemma 3 27B IT
  • open weightsNoYesEnabled
  • knowledge—2024-08
repriced
GLM-4.6

via ofox

What changed

  • Input price$0.4/M$0.6/M▲ 50.0%
  • Output price$1.90/M$2.20/M▲ 15.8%
new model
gpt-oss-20b (GovCloud)

via amazon-bedrock

What changed

First observed in the model catalog

new model
GLM-5V Turbo (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

new model
GPT Terra Latest

via nano-gpt

What changed

First observed in the model catalog

capability
GPT OSS Safeguard 120B

via amazon-bedrock

What changed

  • reasoningNoYesEnabled
new model
Gemma 3 12B IT

via deepinfra

What changed

First observed in the model catalog

new model
Schematron V2 Small

via openrouter

What changed

First observed in the model catalog

context
Llama 4 Maverick 17B Instruct

via amazon-bedrock

What changed

  • descriptionOpen multimodal Llama model for strong reasoning and fast responsesOpen multimodal Llama for strong reasoning with efficient everyday serving
  • Output limit16K8K−50%
new model
GPT-5.6 Terra (India)

via amazon-bedrock

What changed

First observed in the model catalog

new model
Hy4 preview

via llmgateway

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Output limit384K393K1×
  • Input price$1.05/M$0.578/M▼ 44.9%
  • Output price$3.15/M$1.73/M▼ 44.9%
  • Cache read$0.035/M$0.018/M▼ 47.4%
capability
GPT-5.6 Luna (Global)

via amazon-bedrock

What changed

  • structured outputNoYesEnabled
capability
Google Gemma 3 27B Instruct

via venice

What changed

  • knowledge—2024-08
new model
Hy-MT2 Plus (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Pixtral Large (25.02)

via amazon-bedrock

What changed

  • familymistralpixtral
  • attachmentNoYesEnabled
capability
Llama 3.3 70B Instruct

via amazon-bedrock

What changed

  • descriptionOpen Llama instruction model for multilingual chat, reasoning, and codingPopular open Llama workhorse for multilingual chat, coding, and self-hosting
new model
Fugu Max

via llmgateway

What changed

First observed in the model catalog

capability
Voxtral Small 24B 2507

via openrouter

What changed

  • familymistralvoxtral
  • release date2025-10-302025-07-15
  • last updated2025-10-302025-07-15
repriced
GLM-5

via hyper

What changed

  • Output price$2.78/M$2.75/M▼ 1.1%
repriced
Qwen3-Coder 480B-A35B Instruct

via amazon-bedrock

What changed

  • nameQwen3 Coder 480B A35B InstructQwen3-Coder 480B-A35B Instruct
  • descriptionQwen coding model for software agents, repository edits, and code reasoningOpen Qwen coding heavyweight for repository reasoning and agentic engineering
  • knowledge2024-042025-04
  • release date2025-09-182025-07-23
  • +1 more changes
repriced
DeepSeek: DeepSeek V4 Flash Latest

via kilo

What changed

  • nameDeepSeek V4 Flash LatestDeepSeek: DeepSeek V4 Flash Latest
  • Output limit393K131K−67%
  • Input price$0.05/M$0.035/M▼ 29.6%
  • Output price$0.16/M$0.106/M▼ 34.0%
  • +1 more changes
repriced
Qwen3.8 Flash

via ofox

What changed

  • Input price$0.15/M$0.11/M▼ 26.7%
  • Output price$0.47/M$0.39/M▼ 17.0%
  • Cache read$0.016/M$0.011/M▼ 31.3%
  • cache write$0.2/M$0.14/M▼ 30.0%
new model
Kimi K2.7 Code (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

context
Kimi K2 0905

via openrouter

What changed

  • Output limit100K98K−2%
repriced
Nemotron 3 Super 120B A12B

via kilo

What changed

  • structured outputNoYesEnabled
  • Input price$0.085/M$0.08/M▼ 5.9%
  • Output price$0.4/M$0.45/M▲ 12.5%
repriced
GLM-5.2

via openrouter

What changed

  • Output limit131K182K1.4×
  • Input price$0.966/M$0.6/M▼ 37.9%
  • Output price$3.04/M$2/M▼ 34.1%
  • Cache read$0.193/M$0.15/M▼ 22.4%
new model
GPT Luna Latest

via openrouter

What changed

First observed in the model catalog

new model
Mistral Medium 3

via pioneer

What changed

First observed in the model catalog

new model
GPT Sol Latest

via openrouter

What changed

First observed in the model catalog

new model
Gemma 3 27B IT

via merge-gateway

What changed

First observed in the model catalog

removed
OpenAI GPT Latest

via kilo

What changed

Removed from the model catalog

removed
thinkingcap-qwen3.6-27b@eu

via requesty

What changed

Removed from the model catalog

new model
Gemma 3 12B IT (Deep Infra)

via edenai

What changed

First observed in the model catalog

removed
GLM-5.2

via nan

What changed

Removed from the model catalog

capability
MiniMax-M2.1

via amazon-bedrock

What changed

  • nameMiniMax M2.1MiniMax-M2.1
  • descriptionMiniMax model for chat, coding, office work, and agentic tasksEarlier MiniMax agent model for practical coding and productivity tasks
  • structured outputNoYesEnabled
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.404/M$0.396/M▼ 2.0%
  • Output price$1.50/M$1.46/M▼ 2.1%
  • Cache read$0.202/M$0.198/M▼ 2.0%
context
NVIDIA Nemotron Nano 12B v2 VL BF16

via amazon-bedrock

What changed

  • attachmentNoYesEnabled
  • open weightsNoYesEnabled
  • release date2024-12-012025-10-28
  • last updated2024-12-012025-10-28
  • +1 more changes
context
Gemma 3 4B IT

via amazon-bedrock

What changed

  • descriptionOpen Gemma instruction model for efficient chat and self-hosted deploymentsOpen multimodal Gemma instruction model for efficient text generation and image understanding
  • attachmentNoYesEnabled
  • tool callYesNoRemoved
  • open weightsNoYesEnabled
  • +4 more changes
context
NVIDIA Nemotron Nano 9B v2

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
  • release date2024-12-012025-08-18
  • last updated2024-12-012025-08-18
  • Context128K131K1×
  • +1 more changes
context
Kimi K2 Thinking

via openrouter

What changed

  • Output limit100K98K−2%
capability
Nova 2 Lite (US)

via amazon-bedrock

What changed

  • last updated2025-12-022025-12-01
capability
Gemini Flash Latest

via openrouter

What changed

  • nameGoogle Gemini Flash LatestGemini Flash Latest
capability
Devstral 2 123B

via amazon-bedrock

What changed

  • descriptionMistral coding agent model for repository tasks and software engineering workflowsMistral's coding-agent model for repository work, terminal tasks, and software fixes
  • knowledge—2025-12
  • release date2026-02-172025-12-09
  • last updated2026-02-172025-12-09
new model
Llama 3.1 405B

via nano-gpt

What changed

First observed in the model catalog

repriced
Qwen3.6 27B

via ofox

What changed

  • Input price$0.6/M$0.43/M▼ 28.3%
  • Output price$3.60/M$2.57/M▼ 28.6%
capability
DeepSeek-R1 (US)

via amazon-bedrock

What changed

  • tool callYesNoRemoved
new model
Voxtral Small 24B 2507 (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

context
DeepSeek V4 Flash (Consensus Protocol)

via llmgateway-providers

What changed

  • Context1M1.05M1.1×
new model
Fugu Max

via openrouter

What changed

First observed in the model catalog

capability
DeepSeek-V3.1

via amazon-bedrock

What changed

  • descriptionDeepSeek chat model for instruction following, coding, and analysisHybrid-reasoning DeepSeek model with thinking and non-thinking modes
  • knowledge2024-07—
  • release date2025-09-182025-08-21
repriced
Qwen-VL Max

via ofox

What changed

  • Cache read$0.046/M$0.023/M▼ 50.0%
capability
DeepSeek-R1

via vercel

What changed

  • descriptionDeepSeek reasoning model for multi-step analysis, math, coding, and toolsClassic open reasoning model for transparent math, coding, and deliberate problem solving
  • tool callYesNoRemoved
  • open weightsNoYesEnabled
capability
Gemma 3 4B IT

via kilo

What changed

  • nameGoogle: Gemma 3 4BGemma 3 4B IT
  • open weightsNoYesEnabled
  • knowledge—2024-08
  • release date2025-03-132025-03-12
  • +1 more changes
capability
Mistral Large 3

via llmgateway

What changed

  • release date2024-11-012025-12-02
repriced
Voxtral Small 24B 2507

via amazon-bedrock

What changed

  • descriptionEfficient Mistral model for fast chat, extraction, and production assistantsOpen audio-language model for speech transcription, audio understanding, and voice-driven tool use
  • familymistralvoxtral
  • release date2025-07-012025-07-15
  • last updated2025-07-012025-07-15
  • +3 more changes
capability
Gemma 3 27B IT

via cortecs

What changed

  • namegemma-3-27b-itGemma 3 27B IT
  • family—gemma
  • open weightsNoYesEnabled
  • knowledge—2024-08
new model
DeepSeek V4.1 Flash

via aihubmix

What changed

First observed in the model catalog

repriced
GPT-5.6 Sol (EU)

via requesty

What changed

  • Input price$5.50/M$4.40/M▼ 20.0%
  • Output price$33/M$22/M▼ 33.3%
  • Cache read$0.55/M$0.44/M▼ 20.0%
capability
DeepSeek V3.2

via amazon-bedrock

What changed

  • nameDeepSeek-V3.2DeepSeek V3.2
  • descriptionDeepSeek chat model for instruction following, coding, and analysisHybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use
  • release date2026-02-062025-12-01
capability
Llama-3.3-70B-Instruct

via pioneer

What changed

  • attachmentYesNoRemoved
context
GLM-5

via amazon-bedrock

What changed

  • descriptionFlagship GLM model for hybrid reasoning, coding, and agentic engineeringGeneral GLM flagship for coding, analysis, and tool-heavy engineering workflows
  • release date2026-03-182026-02-12
  • Output limit101K131K1.3×
new model
Gemma 3 4B IT

via huggingface

What changed

First observed in the model catalog

capability
GPT Mini Latest

via openrouter

What changed

  • nameOpenAI GPT Mini LatestGPT Mini Latest
new model
Hy4 Preview (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Palmyra X4 (US)

via amazon-bedrock

What changed

  • descriptionReasoning model for deliberate analysis, multi-step problem solving, and tool useEnterprise language model for workflow automation, coding, data analysis, and tool use
  • release date2025-04-282024-10-09
repriced
Llama 4 Maverick 17B Instruct

via hyper

What changed

  • Input price$0.274/M$0.255/M▼ 6.9%
  • Output price$0.899/M$0.837/M▼ 7.0%
  • Cache read$0.137/M$0.128/M▼ 6.9%
new model
OpenAI: GPT Luna Latest

via kilo

What changed

First observed in the model catalog

repriced
Qwen Turbo

via ofox

What changed

  • Input price$0.05/M$0.043/M▼ 14.0%
new model
gpt-oss-120b (GovCloud)

via amazon-bedrock

What changed

First observed in the model catalog

capability
Mistral Large 3

via edenai

What changed

  • release date2024-11-012025-12-02
repriced
Gemini 2.5 Flash-Lite

via ofox

What changed

  • Cache read$0.025/M$0.01/M▼ 60.0%
repriced
GLM-4.7-FlashX

via ofox

What changed

  • Output price$0.43/M$0.4/M▼ 7.0%
  • Cache read$0.015/M$0.01/M▼ 33.3%
new model
Gemma 3 27B IT (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

repriced
DeepSeek V4.1 Flash

via openrouter

What changed

  • Input price$0.3/M$0.15/M▼ 50.0%
  • Output price$1.20/M$0.6/M▼ 50.0%
  • Cache read$0.0060/M$0.0030/M▼ 50.0%
new model
Gemma 3 4B IT (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

capability
MiniMax-M2

via amazon-bedrock

What changed

  • nameMiniMax M2MiniMax-M2
  • descriptionMiniMax model for chat, coding, office work, and agentic tasksEfficient open MiniMax model built for coding agents and tool-heavy workflows
  • structured outputNoYesEnabled
repriced
Gemma 4 26B A4B IT

via openrouter

What changed

  • Output limit33K236K7.2×
  • Input price$0.042/M$0.09/M▲ 114.3%
  • Output price$0.22/M$0.3/M▲ 36.4%
  • Cache read—$0.05/M
new model
GLM-5.1 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Output limit384K393K1×
  • Input price$0.955/M$1.60/M▲ 67.5%
  • Output price$1.91/M$3.20/M▲ 67.5%
  • Cache read$0.08/M$0.135/M▲ 69.6%
new model
Gemma 4 31B IT

via amazon-bedrock

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash (Together AI)

via edenai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via above

What changed

  • descriptionOfficial DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decodingDeepSeek V4.1 Flash model for reasoning and agentic coding
  • attachmentNoYesEnabled
  • release date2026-07-312026-09-10
  • last updated2026-07-312026-09-10
  • +5 more changes
new model
DeepSeek V4.1 Flash (Fireworks AI)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.44/M$0.22/M▼ 50.0%
  • Output price$1.32/M$0.66/M▼ 50.0%
  • Cache read$0.014/M$0.0070/M▼ 50.0%
context
MoonshotAI: Kimi K2 0711

via kilo

What changed

  • Output limit100K98K−2%
new model
Gemma 3 27B IT (Deep Infra)

via edenai

What changed

First observed in the model catalog

capability
Mistral Large 3

via kilo

What changed

  • release date2024-11-012025-12-02
context
GLM-5.3

via kilo

What changed

  • Output limit944K131K−86%
repriced
Llama-3.3-70B-Instruct (IONOS)

via edenai

What changed

  • Input price$0.755/M$0.753/M▼ 0.2%
  • Output price$0.755/M$0.753/M▼ 0.2%
capability
Voxtral Small 24B 2507

via kilo

What changed

  • nameMistral: Voxtral Small 24B 2507Voxtral Small 24B 2507
  • familymistralvoxtral
  • open weightsNoYesEnabled
  • release date2025-10-302025-07-15
  • +1 more changes
capability
Mistral Large 3 675B Instruct 2512

via fireworks-ai

What changed

  • release date2024-11-012025-12-02
capability
Qwen3 235B-A22B Instruct 2507

via amazon-bedrock

What changed

  • nameQwen3 235B A22B 2507Qwen3 235B-A22B Instruct 2507
  • descriptionQwen instruction model for multilingual chat, reasoning, and tool useUpdated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use
  • knowledge2024-04—
  • release date2025-09-182025-07-21
capability
Nova 2 Lite (Global)

via amazon-bedrock

What changed

  • last updated2025-12-022025-12-01
context
Palmyra X5

via amazon-bedrock

What changed

  • Input limit—1.04M
new model
MiMo V2.5 Pro (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Ministral 3 3B

via amazon-bedrock

What changed

  • descriptionCompact Mistral model for edge, latency-sensitive, and cost-efficient workloadsCompact open vision-language model for edge deployment, instruction following, and tool use
  • attachmentNoYesEnabled
capability
Llama-3.3-70B-Instruct

via crusoe

What changed

  • attachmentYesNoRemoved
new model
Muse Spark 1.3 Contributor

via bothub

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731 (Scaleway)

via edenai

What changed

  • Input price$0.465/M$0.464/M▼ 0.2%
  • Output price$0.929/M$0.927/M▼ 0.2%
new model
Kimi K3

via volcengine-coding-plan

What changed

First observed in the model catalog

new model
Gemma 3 12B IT (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

new model
Gemma 3 12B IT (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

new model
GLM-5.2 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Gemma 3 12B IT

via huggingface

What changed

First observed in the model catalog

new model
Fugu Max

via nano-gpt

What changed

First observed in the model catalog

capability
Google: Gemini Flash Latest

via kilo

What changed

  • nameGoogle Gemini Flash LatestGoogle: Gemini Flash Latest
new model
GPT OSS Safeguard 20B

via merge-gateway

What changed

First observed in the model catalog

capability
Anthropic: Claude Sonnet Latest

via kilo

What changed

  • nameAnthropic Claude Sonnet LatestAnthropic: Claude Sonnet Latest
repriced
DeepSeek V4.1 Flash

via nano-gpt

What changed

  • Input price$0.3/M$0.15/M▼ 50.0%
  • Output price$1.20/M$0.6/M▼ 50.0%
  • Cache read$0.0060/M$0.0030/M▼ 50.0%
new model
Magistral Medium (latest)

via pioneer

What changed

First observed in the model catalog

new model
Devstral Small 2

via pioneer

What changed

First observed in the model catalog

capability
GLM-4.7

via amazon-bedrock

What changed

  • descriptionFlagship GLM model for hybrid reasoning, coding, and agentic engineeringMature GLM model for dependable coding, reasoning, and structured agent tasks
context
DeepSeek V4 Pro

via kilo

What changed

  • Context1.02M1.05M1×
  • Output limit384K393K1×
capability
Google: Gemini Pro Latest

via kilo

What changed

  • nameGoogle Gemini Pro LatestGoogle: Gemini Pro Latest
new model
GLM-5-Turbo

via llmgateway

What changed

First observed in the model catalog

new model
Kimi K2.6 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Nemotron 3.5 Lightning 30B A3B

via kilo

What changed

  • Input price$0.08/M$0.065/M▼ 18.8%
  • Output price$0.2/M$0.18/M▼ 10.0%
  • Cache read$0.04/M—
new model
Nvidia Nemotron 3.5 Lightning TEE

via nano-gpt

What changed

First observed in the model catalog

capability
Mistral Large 3

via cortecs

What changed

  • release date2024-11-012025-12-02
new model
DeepSeek V4 Pro (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Anthropic: Claude Haiku Latest

via kilo

What changed

  • nameAnthropic Claude Haiku LatestAnthropic: Claude Haiku Latest
context
Qwen3 32B

via amazon-bedrock

What changed

  • nameQwen3 32B (dense)Qwen3 32B
  • descriptionQwen instruction model for multilingual chat, reasoning, and tool useDense open Qwen model for self-hosted chat, reasoning, and coding
  • knowledge2024-042025-04
  • release date2025-09-182025-04
  • +1 more changes
new model
GLM-5 Turbo (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.02M1.05M1×
  • Output limit384K393K1×
context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
capability
GPT-5.6 Luna (US)

via amazon-bedrock

What changed

  • structured outputNoYesEnabled
new model
Inference.net: Schematron V2 Turbo

via kilo

What changed

First observed in the model catalog

new model
GLM-5.3

via nan

What changed

First observed in the model catalog

context
Llama 4 Maverick 17B Instruct (US)

via amazon-bedrock

What changed

  • Output limit16K8K−50%
new model
Gemma 4 26B A4B IT

via amazon-bedrock

What changed

First observed in the model catalog

capability
Magistral Small 1.2

via amazon-bedrock

What changed

  • descriptionMistral reasoning model for transparent analysis, math, and complex decisionsOpen multimodal reasoning model for transparent analysis of text and images
  • attachmentNoYesEnabled
  • release date2025-12-022025-09-18
  • last updated2025-12-022025-09-18
repriced
Qwen3.8 Max 0902

via ofox

What changed

  • Input price$2/M$1.71/M▼ 14.5%
  • Output price$6/M$5.14/M▼ 14.3%
  • Cache read$0.25/M$0.17/M▼ 32.0%
  • cache write$2.50/M$2.14/M▼ 14.4%
new model
Gemma 3 12B

via merge-gateway

What changed

First observed in the model catalog

context
GLM-5.2

via kilo

What changed

  • Context1M203K−80%
  • Output limit131K182K1.4×
new model
DeepSeek V4.1 Flash (EU)

via requesty

What changed

First observed in the model catalog

capability
GPT OSS Safeguard 20B

via openrouter

What changed

  • namegpt-oss-safeguard-20bGPT OSS Safeguard 20B
repriced
DeepSeek V4 Flash 0731

via cortecs

What changed

  • Input price$0.13/M$0.09/M▼ 30.8%
  • Output price$0.28/M$0.17/M▼ 39.3%
  • Cache read$0.03/M$0.014/M▼ 53.3%
repriced
Qwen3 30B A3B Instruct 2507

via openrouter

What changed

  • Output limit236K32K−86%
  • Input price$0.09/M$0.048/M▼ 46.5%
  • Output price$0.3/M$0.193/M▼ 35.6%
removed
OpenAI GPT Latest

via openrouter

What changed

Removed from the model catalog

new model
Gemma 3 27B IT

via deepinfra

What changed

First observed in the model catalog

repriced
Hy3

via llmgateway

What changed

  • Input price$0.14/M$0.132/M▼ 5.7%
  • Output price$0.58/M$0.528/M▼ 9.0%
  • Cache read$0.035/M$0.033/M▼ 5.7%
new model
Sakana: Fugu Ultra v2

via kilo

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via togetherai

What changed

First observed in the model catalog

capability
Palmyra X4

via amazon-bedrock

What changed

  • descriptionReasoning model for deliberate analysis, multi-step problem solving, and tool useEnterprise language model for workflow automation, coding, data analysis, and tool use
  • release date2025-04-282024-10-09
context
Kimi K2.5

via amazon-bedrock

What changed

  • descriptionKimi multimodal agent model for visual understanding, coding, and planningEarlier Kimi frontier model for long-context agents, coding, and multimodal work
  • familykimikimi-k2
  • attachmentNoYesEnabled
  • knowledge—2025-01
  • +2 more changes
repriced
MoonshotAI: Kimi Latest

via kilo

What changed

  • nameMoonshotAI Kimi LatestMoonshotAI: Kimi Latest
  • Input price$2.34/M$2.10/M▼ 10.3%
  • Output price$11.70/M$10.95/M▼ 6.4%
  • Cache read$0.261/M—
new model
GPT-6 Astra (AWS Mantle)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Gemma 3 27B IT

via huggingface

What changed

First observed in the model catalog

capability
gpt-oss-20b

via amazon-bedrock

What changed

  • descriptionOpen-weight GPT model for self-hosted reasoning and instruction-following workloadsOpen GPT reasoning model for self-hosted agents and controllable deployments
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context1.05M262K−75%
  • Output limit944K236K−75%
  • Input price$1/M$0.936/M▼ 6.4%
  • Output price$3.41/M$3.17/M▼ 7.1%
  • +1 more changes
new model
OpenAI: GPT Terra Latest

via kilo

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via greenpt

What changed

First observed in the model catalog

repriced
Qwen3.8 Max

via ofox

What changed

  • Input price$2/M$1.71/M▼ 14.5%
  • Output price$6/M$5.14/M▼ 14.3%
  • Cache read$0.25/M$0.17/M▼ 32.0%
  • cache write$2.50/M$2.14/M▼ 14.4%
new model
Gemma 3 27B IT (Scaleway)

via edenai

What changed

First observed in the model catalog

repriced
Fugu Ultra v1.1

via nano-gpt

What changed

  • Input price$5.25/M$5/M▼ 4.8%
  • Output price$31.50/M$30/M▼ 4.8%
  • Cache read$0.525/M$0.5/M▼ 4.8%
context
NVIDIA Nemotron Nano 3 30B

via amazon-bedrock

What changed

  • release date2025-12-232025-12-15
  • Context128K262K2×
  • Output limit4K8K2×
capability
Gemma 3 4B IT

via openrouter

What changed

  • nameGemma 3 4BGemma 3 4B IT
  • knowledge2024-08-312024-08
  • release date2025-03-132025-03-12
  • last updated2025-03-132025-03-12
new model
Fugu Ultra v2.0

via empiriolabs

What changed

First observed in the model catalog

new model
OpenAI: GPT Astra Latest ($$$$)

via kilo

What changed

First observed in the model catalog

repriced
Qwen3.8 27B

via cortecs

What changed

  • Input price$0.334/M$0.1/M▼ 70.1%
  • Output price$2.45/M$0.4/M▼ 83.7%
  • Cache read$0.111/M$0.04/M▼ 64.0%
new model
GPT Astra Latest

via openrouter

What changed

First observed in the model catalog

repriced
GPT OSS 120B (IONOS)

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.2%
  • Output price$0.755/M$0.753/M▼ 0.2%
capability
Llama 3.1 8B Instruct

via amazon-bedrock

What changed

  • descriptionOpen Llama instruction model for multilingual chat, reasoning, and codingCompact open Llama model for lightweight chat, drafting, and self-hosting
context
Inkling

via openrouter

What changed

  • Output limit33K472K14.4×
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.178/M$0.168/M▼ 5.6%
  • Output price$0.68/M$0.66/M▼ 2.9%
  • Cache read$0.089/M$0.084/M▼ 5.6%
capability
Nemotron 3 Super 120B A12B

via openrouter

What changed

  • structured outputNoYesEnabled
removed
Thinking Machines: Inkling (free)

via kilo

What changed

Removed from the model catalog

context
Voxtral Mini 3B 2507

via amazon-bedrock

What changed

  • descriptionEfficient Mistral model for fast chat, extraction, and production assistantsOpen audio-language model for speech transcription, audio understanding, and voice-driven tool use
  • familymistralvoxtral
  • attachmentNoYesEnabled
  • open weightsNoYesEnabled
  • +4 more changes
repriced
GPT OSS 120B (Scaleway)

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.2%
  • Output price$0.697/M$0.696/M▼ 0.2%
repriced
Fugu Ultra

via nano-gpt

What changed

  • Input price$5.25/M$5/M▼ 4.8%
  • Output price$31.50/M$30/M▼ 4.8%
  • Cache read$0.525/M$0.5/M▼ 4.8%
new model
Hy3 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

new model
GPT Sol Latest

via nano-gpt

What changed

First observed in the model catalog

new model
GPT OSS Safeguard 20B (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit393K131K−67%
  • Input price$0.05/M$0.035/M▼ 29.6%
  • Output price$0.16/M$0.106/M▼ 34.0%
  • Cache read$0.013/M$0.0011/M▼ 91.4%
capability
MiniMax-M2.5

via amazon-bedrock

What changed

  • nameMiniMax M2.5MiniMax-M2.5
  • descriptionMiniMax model for chat, coding, office work, and agentic tasksPrior MiniMax coding model for agent workflows, office edits, and automation
  • structured outputNoYesEnabled
  • release date2026-03-182026-02-12
capability
gpt-oss-20b

via amazon-bedrock

What changed

  • descriptionOpen-weight GPT model for self-hosted reasoning and instruction-following workloadsOpen GPT reasoning model for self-hosted agents and controllable deployments
new model
Ministral 14B

via pioneer

What changed

First observed in the model catalog

new model
Fugu Ultra v2.0

via llmgateway

What changed

First observed in the model catalog

context
Qwen: Qwen3 30B A3B Instruct 2507

via kilo

What changed

  • Context262K128K−51%
  • Output limit236K32K−86%
new model
Schematron V2 Turbo

via openrouter

What changed

First observed in the model catalog

new model
Voxtral Mini 3B 2507 (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

new model
Fugu Ultra v2

via openrouter

What changed

First observed in the model catalog

context
Qwen3.8 27B

via kilo

What changed

  • Context1M262K−74%
context
Kimi K2 Thinking

via kilo

What changed

  • Output limit100K98K−2%
repriced
GLM5.3

via digitalocean

What changed

  • Output limit1.05M128K−88%
  • Input price$1.40/M$0.95/M▼ 32.1%
  • Output price$4.40/M$3.40/M▼ 22.7%
  • Cache read$0.26/M$0.2/M▼ 23.1%
context
Inkling

via kilo

What changed

  • Context1.05M524K−50%
  • Output limit33K472K14.4×
repriced
Qwen3.7 Max

via ofox

What changed

  • Input price$2.50/M$1.71/M▼ 31.6%
  • Output price$7.50/M$5.14/M▼ 31.5%
  • Cache read$0.5/M$0.17/M▼ 66.0%
  • cache write$3.13/M$2.14/M▼ 31.5%
repriced
DeepSeek V4 Pro 0813 (Alibaba)

via edenai

What changed

  • Input price$0.581/M$1.12/M▲ 93.2%
  • Output price$1.74/M$3.37/M▲ 93.2%
new model
Gemma 3 4B IT (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

new model
Voxtral Small 24B 2507 (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

capability
Nova 2 Lite

via amazon-bedrock

What changed

  • last updated2025-12-022025-12-01
capability
DeepSeek-R1

via openrouter

What changed

  • structured outputYesNoRemoved
capability
Llama-3.3-70B-Instruct

via snowflake-cortex

What changed

  • attachmentYesNoRemoved
new model
Mistral Small 24B

via nano-gpt

What changed

First observed in the model catalog

new model
Ministral 3 14B (Infomaniak)

via edenai

What changed

First observed in the model catalog

capability
gpt-oss-120b

via amazon-bedrock

What changed

  • descriptionOpen-weight GPT model for self-hosted reasoning and instruction-following workloadsOpen GPT reasoning model for self-hosted agents and controllable deployments
capability
Llama-3.3-70B-Instruct

via watsonx

What changed

  • attachmentYesNoRemoved
new model
DeepSeek V4.1 Flash

via cline-pass

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.122/M$0.102/M▼ 16.4%
  • Output price$0.42/M$0.356/M▼ 15.2%
  • Cache read$0.061/M$0.051/M▼ 16.4%
new model
Gemma 3 4B IT (Deep Infra)

via edenai

What changed

First observed in the model catalog

capability
Llama-3.3-70B-Instruct

via evroc

What changed

  • attachmentYesNoRemoved
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.065/M$0.04/M▼ 38.5%
  • Output price$0.18/M$0.08/M▼ 55.6%
  • Cache read$0.016/M$0.0080/M▼ 50.0%
new model
GLM-5 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
GPT OSS Safeguard 20B

via nano-gpt

What changed

  • release date2026-02-232025-10-29
context
Llama-3.1-70B-Instruct

via kilo

What changed

  • Output limit8K16K2×
new model
GPT OSS Safeguard 20B (Groq)

via edenai

What changed

First observed in the model catalog

capability
Mistral Large 3

via opper

What changed

  • release date2024-11-012025-12-02
new model
GPT OSS Safeguard 20B (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

context
Gemma 4 26B A4B IT

via kilo

What changed

  • Context131K262K2×
  • Output limit33K236K7.2×
new model
OpenAI: GPT Sol Latest

via kilo

What changed

First observed in the model catalog

repriced
Kimi Latest

via openrouter

What changed

  • nameMoonshotAI Kimi LatestKimi Latest
  • Input price$2.34/M$2.10/M▼ 10.3%
  • Output price$11.70/M$10.95/M▼ 6.4%
  • Cache read$0.261/M—
repriced
Llama-3.1-70B-Instruct

via openrouter

What changed

  • Output limit8K16K2×
  • Input price$0.72/M$0.4/M▼ 44.4%
  • Output price$0.72/M$0.4/M▼ 44.4%
new model
Sakana: Fugu Max

via kilo

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via baseten

What changed

First observed in the model catalog

repriced
GLM-5.3

via openrouter

What changed

  • Output limit944K131K−86%
  • Input price$1.40/M$1.09/M▼ 22.0%
  • Output price$4.40/M$3.43/M▼ 22.0%
  • Cache read$0.26/M$0.203/M▼ 22.0%
new model
DeepSeek V4.1 Flash (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4.1 Flash

via requesty

What changed

  • Input price$0.3/M$0.5/M▲ 66.7%
  • Output price$1.20/M$1.50/M▲ 25.0%
  • Cache read$0.0060/M$0.05/M▲ 733.3%
repriced
GPT-5.4 Pro

via edenai

What changed

  • Cache read$3/M—
  • tiers[object Object][object Object]
repriced
GLM-5.3-Flash

via cortecs

What changed

  • Input price$0.201/M$0.1/M▼ 50.2%
  • Output price$0.5/M$0.35/M▼ 30.0%
  • Cache read$0.05/M$0.018/M▼ 64.0%
repriced
Qwen3.8 27B

via openrouter

What changed

  • Input price$0.42/M$0.214/M▼ 49.0%
  • Output price$3/M$2.55/M▼ 15.0%
  • Cache read$0.085/M$0.15/M▲ 76.5%
repriced
DeepSeek V3 0324

via openrouter

What changed

  • Input price$0.29/M$0.25/M▼ 13.8%
  • Output price$1.14/M$1/M▼ 12.3%
  • Cache read$0.11/M—
capability
Mistral Large 3 (Mistral AI)

via llmgateway-providers

What changed

  • release date2024-11-012025-12-02
context
Fugu Ultra

via merge-gateway

What changed

  • Output limit250K128K−49%
new model
GPT Luna Latest

via nano-gpt

What changed

First observed in the model catalog

new model
Fugu Ultra v2.0 (Sakana AI)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via merge-gateway

What changed

  • Context1M1.05M1×
  • Input price$0.22/M$0.035/M▼ 84.1%
  • Output price$0.66/M$0.07/M▼ 89.4%
repriced
Gemma 3 12B IT

via amazon-bedrock

What changed

  • nameGoogle Gemma 3 12BGemma 3 12B IT
  • descriptionOpen Gemma instruction model for efficient chat and self-hosted deploymentsOpen multimodal Gemma instruction model for multilingual text generation and image understanding
  • attachmentNoYesEnabled
  • open weightsNoYesEnabled
  • +5 more changes
repriced
Qwen3-Next 80B-A3B Instruct

via amazon-bedrock

What changed

  • nameQwen/Qwen3-Next-80B-A3B-InstructQwen3-Next 80B-A3B Instruct
  • open weightsNoYesEnabled
  • knowledge—2025-04
  • release date2025-09-182025-09-11
  • +3 more changes
context
Qwen: Qwen3 235B A22B Instruct 2507

via kilo

What changed

  • Output limit16K236K14.4×
capability
Llama-3.3-70B-Instruct

via llmgateway

What changed

  • attachmentYesNoRemoved
repriced
Qwen3.8 27B

via ofox

What changed

  • Input price$0.45/M$0.5/M▲ 11.1%
  • Output price$3.20/M$1.71/M▼ 46.6%
  • Cache read$0.05/M$0.043/M▼ 14.0%
  • cache write$0.563/M$0.63/M▲ 12.0%
repriced
GPT-5.5 Pro

via edenai

What changed

  • Cache read$3/M—
  • tiers[object Object][object Object]
repriced
DeepSeek V4 Pro 0423

via ofox

What changed

  • Cache read$0.044/M$0.15/M▲ 240.9%
repriced
DeepSeek: DeepSeek V3 0324

via kilo

What changed

  • Input price$0.29/M$0.25/M▼ 13.8%
  • Output price$1.14/M$1/M▼ 12.3%
  • Cache read$0.11/M—
capability
Mistral Large 3

via openrouter

What changed

  • release date2024-11-012025-12-02
capability
Nova 2 Lite (EU)

via amazon-bedrock

What changed

  • last updated2025-12-022025-12-01
capability
DeepSeek-R1

via kilo

What changed

  • structured outputYesNoRemoved
context
MoonshotAI: Kimi K2 0905

via kilo

What changed

  • Output limit100K98K−2%
new model
DeepSeek V4 Flash

via agentrouter

What changed

First observed in the model catalog

capability
DeepSeek-R1

via amazon-bedrock

What changed

  • descriptionDeepSeek reasoning model for multi-step analysis, math, coding, and toolsClassic open reasoning model for transparent math, coding, and deliberate problem solving
  • tool callYesNoRemoved
  • open weightsNoYesEnabled
repriced
DeepSeek V4.1 Flash Thinking

via nano-gpt

What changed

  • Input price$0.3/M$0.15/M▼ 50.0%
  • Output price$1.20/M$0.6/M▼ 50.0%
  • Cache read$0.0060/M$0.0030/M▼ 50.0%
repriced
MiniMax-M2.5

via openrouter

What changed

  • Output limit131K128K−2%
  • Input price$0.3/M$0.27/M▼ 10.0%
  • Output price$1.20/M$1.08/M▼ 10.0%
  • Cache read$0.03/M$0.027/M▼ 10.0%
capability
Kimi K2 Thinking

via amazon-bedrock

What changed

  • descriptionKimi reasoning model for long-horizon research, planning, and tool useThinking Kimi model for slower research passes, planning, and hard technical questions
  • knowledge—2024-08
  • release date2025-12-022025-11-06
capability
Nova 2 Lite

via cortecs

What changed

  • last updated2025-12-022025-12-01
new model
DeepSeek V4 Pro 0813

via ollama-cloud

What changed

First observed in the model catalog

repriced
Muse Glimmer 30B

via openrouter

What changed

  • Input price$0.3/M$0.35/M▲ 16.7%
  • Output price$1.10/M$1.50/M▲ 36.4%
new model
GLM-5.3-Flash

via nvidia

What changed

First observed in the model catalog

new model
Inference.net: Schematron V2 Small

via kilo

What changed

First observed in the model catalog

capability
Claude Haiku Latest

via openrouter

What changed

  • nameAnthropic Claude Haiku LatestClaude Haiku Latest
new model
Gemma 4 E2B IT

via amazon-bedrock

What changed

First observed in the model catalog

new model
GPT-5.6 Luna (India)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GLM-5.1

via hyper

What changed

  • Input price$1.33/M$1.32/M▼ 0.6%
  • Output price$4.22/M$4.31/M▲ 2.1%
  • Cache read$0.663/M$0.659/M▼ 0.6%
context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
new model
DeepSeek V4.1 Flash (DeepInfra)

via llmgateway-providers

What changed

First observed in the model catalog

new model
GPT Astra Latest

via nano-gpt

What changed

First observed in the model catalog

capability
Mistral Large 3

via pioneer

What changed

  • release date2024-11-012025-12-02
repriced
Gemma 3 27B IT

via amazon-bedrock

What changed

  • nameGoogle Gemma 3 27B InstructGemma 3 27B IT
  • descriptionOpen Gemma instruction model for efficient chat and self-hosted deploymentsLargest open Gemma 3 instruction model for multilingual text generation and visual understanding
  • tool callYesNoRemoved
  • knowledge2025-072024-08
  • +4 more changes
capability
GLM-4.7-Flash

via amazon-bedrock

What changed

  • descriptionEfficient GLM model for fast reasoning, coding, and agent workflowsBudget GLM lane for fast coding help, routing, and everyday automation
repriced
Qwen3 235B A22B Instruct 2507

via openrouter

What changed

  • Output limit16K236K14.4×
  • Input price$0.22/M$0.087/M▼ 60.2%
  • Output price$0.88/M$0.35/M▼ 60.2%
  • Cache read—$0.018/M
new model
Kimi K2.7 Code Highspeed (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

new model
MiniMax M3 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Gemini Pro Latest

via openrouter

What changed

  • nameGoogle Gemini Pro LatestGemini Pro Latest
new model
GPT Terra Latest

via openrouter

What changed

First observed in the model catalog

new model
Kimi K3 (Runpod)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.089/M$0.049/M▼ 44.7%
  • Output price$0.177/M$0.098/M▼ 44.7%
  • Cache read$0.018/M$0.0098/M▼ 44.7%
removed
thinkingcap-qwen3.6-27b

via requesty

What changed

Removed from the model catalog

capability
Gemma 3 27B IT

via openrouter

What changed

  • nameGemma 3 27BGemma 3 27B IT
  • knowledge2024-08-312024-08
capability
Claude Sonnet Latest

via openrouter

What changed

  • nameAnthropic Claude Sonnet LatestClaude Sonnet Latest
new model
Ling 3.0 Flash VL

via openrouter

What changed

First observed in the model catalog

context
Kimi K2 0711

via openrouter

What changed

  • Output limit100K98K−2%
new model
MiniMax M2.7 (Tencent Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Claude Opus 5

via pioneer

What changed

First observed in the model catalog

new model
Fugu Max (Sakana AI)

via llmgateway-providers

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via ollama-cloud

What changed

First observed in the model catalog

capability
Mistral Small 3.2 24B Instruct

via venice

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
capability
OpenAI: GPT Mini Latest

via kilo

What changed

  • nameOpenAI GPT Mini LatestOpenAI: GPT Mini Latest
new model
Kimi K3 (Alibaba Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

capability
DeepSeek V4.1 Flash

via nan

What changed

  • nameDeepSeek V4 FlashDeepSeek V4.1 Flash
  • descriptionFast DeepSeek V4 lane for economical reasoning, coding, and long-context workDeepSeek V4.1 Flash model for reasoning and agentic coding
  • attachmentNoYesEnabled
  • release date2026-04-242026-09-10
  • +2 more changes
new model
Ministral 3B

via pioneer

What changed

First observed in the model catalog

repriced
Kimi K3

via openrouter

What changed

  • Input price$2.34/M$2.65/M▲ 13.2%
  • Output price$11.70/M$13.28/M▲ 13.5%
  • Cache read$0.261/M$0.303/M▲ 16.0%
new model
Inkling Small

via pioneer

What changed

First observed in the model catalog

new model
LightOnOCR 2

via nano-gpt

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via volcengine-coding-plan

What changed

First observed in the model catalog

new model
Kimi K3

via pioneer

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct (Scaleway)

via edenai

What changed

  • Input price$1.05/M$1.04/M▼ 0.2%
  • Output price$1.05/M$1.04/M▼ 0.2%
capability
GPT OSS Safeguard 20B

via kilo

What changed

  • nameOpenAI: gpt-oss-safeguard-20bGPT OSS Safeguard 20B
  • open weightsNoYesEnabled

172 events
capability
GLM 5.3

via baseten

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
new model
Qwen Max

via ofox

What changed

First observed in the model catalog

new model
Qwen3.5 122B-A10B

via ofox

What changed

First observed in the model catalog

new model
Grok 4.1 Fast

via google-vertex

What changed

First observed in the model catalog

repriced
DeepSeek V4.1 Flash

via nano-gpt

What changed

  • descriptionDeepSeek V4.1 Flash supports text and image input, reasoning, tool calling, and structured output with a 1M-token context window. This is a rate-limited beta with limited capacity, intended for testing rather than production use. Assume prompts and responses are logged by the provider and may be used for model training or service improvement. Do not send sensitive or confidential data.DeepSeek V4.1 Flash model for reasoning and agentic coding
  • familydeepseekdeepseek-flash
  • open weightsNoYesEnabled
  • knowledge—2025-05
  • +5 more changes
repriced
Nova Micro (EU)

via amazon-bedrock

What changed

  • Output limit8K10K1.2×
  • cache write—$0.04/M
context
Llama-3.1-70B-Instruct

via kilo

What changed

  • Output limit16K8K−50%
new model
Qwen3.5 Plus

via ofox

What changed

First observed in the model catalog

repriced
Z.ai: GLM Flash Latest

via kilo

What changed

  • Input price$0.07/M$0.075/M▲ 7.1%
  • Output price$0.233/M$0.25/M▲ 7.2%
  • Cache read$0.014/M$0.015/M▲ 7.1%
new model
Qwen3.8 Max

via ofox

What changed

First observed in the model catalog

new model
Nova Premier (US)

via amazon-bedrock

What changed

First observed in the model catalog

removed
MiMo-V2.5-Pro

via scnet-token-plan

What changed

Removed from the model catalog

removed
DeepSeek V3.2

via scnet-token-plan

What changed

Removed from the model catalog

repriced
GPT OSS 120B (IONOS)

via edenai

What changed

  • Input price$0.175/M$0.174/M▼ 0.3%
  • Output price$0.757/M$0.755/M▼ 0.3%
repriced
Nova Pro (EU)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.92/M
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.18/M$0.178/M▼ 1.1%
  • Output price$0.61/M$0.68/M▲ 11.5%
  • Cache read—$0.089/M
  • cache write$0.09/M—
new model
Gemini 3.5 Flash Lite

via sap-ai-core

What changed

First observed in the model catalog

repriced
Nova Lite (CA)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.064/M
repriced
GLM-5.3-Flash

via openrouter

What changed

  • Input price$0.07/M$0.15/M▲ 114.3%
  • Output price$0.233/M$0.5/M▲ 114.3%
  • Cache read$0.014/M$0.03/M▲ 114.3%
new model
Qwen3 Coder Plus

via ofox

What changed

First observed in the model catalog

repriced
Nova Lite (EU)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.069/M
new model
DeepSeek V4.1 Flash

via vercel

What changed

First observed in the model catalog

removed
DeepSeek V4.1 Flash Beta

via vercel

What changed

Removed from the model catalog

new model
Qwen3.5 35B-A3B

via ofox

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via crossmodel

What changed

First observed in the model catalog

context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
new model
GPT-5.6 Luna

via bothub

What changed

First observed in the model catalog

removed
K2-Horizon-7B

via nano-gpt

What changed

Removed from the model catalog

new model
DeepSeek V4.1 Flash

via requesty

What changed

First observed in the model catalog

new model
Qwen3.6 Max Preview

via ofox

What changed

First observed in the model catalog

repriced
Nova Pro (US)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.8/M
new model
DeepSeek V4 Pro 0813

via bothub

What changed

First observed in the model catalog

context
Nova Lite (US)

via edenai

What changed

  • Output limit8K10K1.2×
repriced
Qwen3-Next 80B-A3B Instruct

via hyper

What changed

  • Cache read—$0.059/M
  • cache write$0.059/M—
repriced
DeepSeek V4 Flash 0731

via nano-gpt

What changed

  • Input price$0.14/M$0.05/M▼ 64.3%
  • Output price$0.28/M$0.16/M▼ 42.9%
  • Cache read$0.014/M$0.013/M▼ 7.1%
new model
Claude Fable 5.1 (AWS Bedrock)

via llmgateway-providers

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via llmgateway

What changed

First observed in the model catalog

removed
DeepSeek V4 Flash Vision Exp

via llmgateway

What changed

Removed from the model catalog

new model
DeepSeek V4.1 Flash

via opencode-go

What changed

First observed in the model catalog

capability
GLM-5.3

via hyper

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
new model
GLM-5.3 Flash (vichar-ai)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Nova Micro (APAC)

via amazon-bedrock

What changed

  • Output limit8K10K1.2×
  • cache write—$0.037/M
repriced
Llama-3.1-70B-Instruct

via openrouter

What changed

  • Output limit16K8K−50%
  • Input price$0.4/M$0.72/M▲ 80.0%
  • Output price$0.4/M$0.72/M▲ 80.0%
context
DeepSeek V4 Flash (Consensus Protocol)

via llmgateway-providers

What changed

  • Context524K1M1.9×
new model
Grok 4.6

via google-vertex

What changed

First observed in the model catalog

new model
DeepSeek V4 Pro 0813 (Nebius)

via edenai

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via huggingface

What changed

First observed in the model catalog

repriced
Kimi K2 Thinking

via hyper

What changed

  • Cache read—$0.3/M
  • cache write$0.3/M—
repriced
Nova 2 Lite (Global)

via amazon-bedrock

What changed

  • knowledge—2025-10
  • release date2025-12-012025-12-02
  • last updated2025-12-012025-12-02
  • Output limit64K66K1×
  • +1 more changes
new model
DeepSeek V4.1 Flash

via fireworks-ai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via opencode-go

What changed

  • Input price$0.22/M$0.15/M▼ 31.8%
  • Output price$0.66/M$0.6/M▼ 9.1%
  • Cache read$0.0070/M$0.0030/M▼ 57.1%
new model
Qwen3.6 27B

via ofox

What changed

First observed in the model catalog

new model
Grok 4.1 Fast (Reasoning)

via google-vertex

What changed

First observed in the model catalog

new model
Qwen3.6 Flash

via ofox

What changed

First observed in the model catalog

removed
Ornith 1.5 9B Thinking

via nano-gpt

What changed

Removed from the model catalog

removed
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

Removed from the model catalog

new model
Mercury 2.5

via inception

What changed

First observed in the model catalog

deprecated
GLM-4.7-Flash

via deepinfra

What changed

  • status—deprecated
repriced
MiMo-V2.5

via deepinfra

What changed

  • Input price$0.4/M$0.14/M▼ 65.0%
  • Output price$2/M$0.28/M▼ 86.0%
  • Cache read$0.08/M$0.0028/M▼ 96.5%
repriced
Nova Micro (US)

via amazon-bedrock

What changed

  • Output limit8K10K1.2×
  • cache write—$0.035/M
repriced
Nova Pro

via vercel

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.8/M
repriced
DeepSeek V4 Flash Latest

via nano-gpt

What changed

  • Input price$0.14/M$0.05/M▼ 64.3%
  • Output price$0.28/M$0.16/M▼ 42.9%
  • Cache read$0.014/M$0.013/M▼ 7.1%
new model
Qwen3.5 Flash

via ofox

What changed

First observed in the model catalog

context
Nova Pro (US)

via edenai

What changed

  • Output limit8K10K1.2×
deprecated
GLM-5

via deepinfra

What changed

  • status—deprecated
repriced
Nova Lite (APAC)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.063/M
new model
Qwen3 Max

via ofox

What changed

First observed in the model catalog

new model
Qwen3.5 397B-A17B

via ofox

What changed

First observed in the model catalog

repriced
GLM-5.3

via ofox

What changed

  • Input price$1.26/M$1.40/M▲ 11.1%
  • Output price$3.96/M$4.40/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
new model
Qwen3.6 Plus

via ofox

What changed

First observed in the model catalog

context
GLM-4.7-Flash

via openrouter

What changed

  • Context203K200K−1%
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.39/M$1.33/M▼ 4.3%
  • Output price$4.36/M$4.22/M▼ 3.1%
  • Cache read—$0.663/M
  • cache write$0.693/M—
repriced
DeepSeek V4 Flash

via deepseek

What changed

  • descriptionOfficial DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decodingDeepSeek V4.1 Flash model for reasoning and agentic coding
  • attachmentNoYesEnabled
  • release date2026-07-312026-09-10
  • last updated2026-07-312026-09-10
  • +5 more changes
new model
Qwen3.8 Max 0902

via ofox

What changed

First observed in the model catalog

removed
DeepSeek V4 Flash Vision Exp Uncensored

via nano-gpt

What changed

Removed from the model catalog

repriced
Gemma 4 31B IT

via llmgateway

What changed

  • Input price$0.102/M$0.1/M▼ 2.0%
  • Output price$0.297/M$0.25/M▼ 15.8%
  • Cache read$0.012/M$0.01/M▼ 16.7%
new model
Qwen-VL Max

via ofox

What changed

First observed in the model catalog

repriced
GPT-OSS 20B TEE

via nano-gpt

What changed

  • Input price$0.2/M$0.04/M▼ 80.0%
  • Output price$0.8/M$0.15/M▼ 81.3%
  • Cache read$0.1/M$0.02/M▼ 80.0%
repriced
DeepSeek V4.1 Flash Thinking

via nano-gpt

What changed

  • descriptionDeepSeek V4.1 Flash supports text and image input, reasoning, tool calling, and structured output with a 1M-token context window. This is a rate-limited beta with limited capacity, intended for testing rather than production use. Assume prompts and responses are logged by the provider and may be used for model training or service improvement. Do not send sensitive or confidential data.DeepSeek V4.1 Flash model for reasoning and agentic coding
  • familydeepseekdeepseek-flash
  • open weightsNoYesEnabled
  • knowledge—2025-05
  • +5 more changes
new model
DeepSeek V4.1 Flash (Deep Infra)

via edenai

What changed

First observed in the model catalog

repriced
MoonshotAI Kimi Latest

via openrouter

What changed

  • Input price$2.40/M$2.34/M▼ 2.5%
  • Output price$12/M$11.70/M▼ 2.5%
  • Cache read$0.24/M$0.261/M▲ 8.8%
new model
DeepSeek V4.1 Flash

via ofox

What changed

First observed in the model catalog

repriced
Qwen3.8 Flash

via nano-gpt

What changed

  • Input price$0.16/M$0.14/M▼ 12.5%
  • Output price$0.47/M$0.42/M▼ 10.6%
new model
DeepSeek V4.1 Flash

via hyper

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash 0423

via ofox

What changed

First observed in the model catalog

removed
DeepSeek V4 Flash (DeepSeek)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Llama 4 Maverick 17B Instruct

via hyper

What changed

  • Cache read—$0.137/M
  • cache write$0.137/M—
repriced
Nova Lite

via vercel

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.06/M
repriced
Nova Pro

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.8/M
repriced
DeepSeek V3.2

via vercel

What changed

  • Input price$0.28/M$0.62/M▲ 121.4%
  • Output price$0.42/M$1.85/M▲ 340.5%
  • Cache read$0.028/M—
repriced
GLM-5.2

via ofox

What changed

  • Input price$0.98/M$1.40/M▲ 42.9%
  • Output price$3.08/M$4.40/M▲ 42.9%
  • Cache read$0.182/M$0.26/M▲ 42.9%
new model
Qwen3.8 Flash

via ofox

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via deepseek

What changed

  • descriptionExperimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic workDeepSeek V4.1 Flash model for reasoning and agentic coding
  • statusbeta—
  • open weightsNoYesEnabled
  • knowledge—2025-05
  • +6 more changes
repriced
Llama-3.3-70B-Instruct (Scaleway)

via edenai

What changed

  • Input price$1.05/M$1.05/M▼ 0.3%
  • Output price$1.05/M$1.05/M▼ 0.3%
new model
DeepSeek V4.1 Flash

via venice

What changed

First observed in the model catalog

repriced
Qwen3-Coder 480B-A35B Instruct

via hyper

What changed

  • Cache read—$0.223/M
  • cache write$0.223/M—
new model
DeepSeek V4 Flash 0731

via bothub

What changed

First observed in the model catalog

new model
Grok 4.3

via google-vertex

What changed

First observed in the model catalog

new model
inclusionAI: Ling 3.0 Flash VL (free)

via kilo

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct (IONOS)

via edenai

What changed

  • Input price$0.757/M$0.755/M▼ 0.3%
  • Output price$0.757/M$0.755/M▼ 0.3%
repriced
Nova Pro (APAC)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.84/M
new model
GLM-5.3 (vichar-ai)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via openrouter

What changed

  • Output limit16K33K2×
  • Input price$0.07/M$0.042/M▼ 40.0%
  • Output price$0.34/M$0.22/M▼ 35.3%
removed
Ornith 1.5 9B

via nano-gpt

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731 (Thinking)

via nano-gpt

What changed

  • Input price$0.14/M$0.05/M▼ 64.3%
  • Output price$0.28/M$0.16/M▼ 42.9%
  • Cache read$0.014/M$0.013/M▼ 7.1%
deprecated
Qwen3.8 Max Preview

via alibaba-token-plan-cn

What changed

  • statusbetadeprecated
repriced
Nova Micro

via vercel

What changed

  • Output limit8K10K1.2×
  • cache write—$0.035/M
repriced
Solar Pro 4

via openrouter

What changed

  • Input price$0.03/M$0.09/M▲ 200.0%
  • Output price$0.12/M$0.36/M▲ 200.0%
  • Cache read$0.0060/M$0.018/M▲ 200.0%
repriced
GLM Flash Latest

via openrouter

What changed

  • Input price$0.07/M$0.075/M▲ 7.1%
  • Output price$0.233/M$0.25/M▲ 7.2%
  • Cache read$0.014/M$0.015/M▲ 7.1%
new model
Qwen Turbo

via ofox

What changed

First observed in the model catalog

capability
Nova 2 Lite

via cortecs

What changed

  • knowledge—2025-10
  • release date2025-12-012025-12-02
  • last updated2025-12-012025-12-02
repriced
Nova Lite

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.06/M
new model
DeepSeek V4.1 Flash (DeepSeek)

via llmgateway-providers

What changed

First observed in the model catalog

context
Nova Pro

via edenai

What changed

  • Output limit8K10K1.2×
new model
DeepSeek V4.1 Flash

via deepseek

What changed

First observed in the model catalog

deprecated
Omen Alpha

via opencode-go

What changed

  • status—deprecated
new model
Gemini 3.8 Flash

via ofox

What changed

First observed in the model catalog

new model
Qwen3.5 27B

via ofox

What changed

First observed in the model catalog

new model
Grok 4.20 (Non-Reasoning)

via google-vertex

What changed

First observed in the model catalog

repriced
GLM-5.3-Flash

via ofox

What changed

  • Input price$0.075/M$0.15/M▲ 100.0%
  • Output price$0.25/M$0.5/M▲ 100.0%
  • Cache read$0.015/M$0.03/M▲ 100.0%
new model
Ling 3.0 Flash VL (free)

via openrouter

What changed

First observed in the model catalog

context
Nova Micro (US)

via edenai

What changed

  • Output limit8K10K1.2×
repriced
Nova Lite (US)

via amazon-bedrock

What changed

  • modalities.inputtext,image,videotext,image,video,pdf
  • Output limit8K10K1.2×
  • cache write—$0.06/M
new model
Qwen3 Coder Next

via ofox

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731 (Scaleway)

via edenai

What changed

  • Input price$0.466/M$0.465/M▼ 0.3%
  • Output price$0.932/M$0.929/M▼ 0.3%
removed
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

Removed from the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • Cache read$0.033/M$0.035/M▲ 6.1%
repriced
Nova 2 Lite (JP)

via amazon-bedrock

What changed

  • knowledge—2025-10
  • release date2025-12-012025-12-02
  • last updated2025-12-012025-12-02
  • Output limit64K66K1×
  • +1 more changes
repriced
Nova 2 Lite

via amazon-bedrock

What changed

  • attachmentNoYesEnabled
  • knowledge—2025-10
  • release date2024-12-012025-12-02
  • last updated2024-12-012025-12-02
  • +5 more changes
new model
Qwen3 Coder Flash

via ofox

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via merge-gateway

What changed

First observed in the model catalog

repriced
GPT OSS 120B (Scaleway)

via edenai

What changed

  • Input price$0.175/M$0.174/M▼ 0.3%
  • Output price$0.699/M$0.697/M▼ 0.3%
new model
GLM-5.3-Flash

via bothub

What changed

First observed in the model catalog

removed
MiMo V2.5 Pro UltraSpeed

via above

What changed

Removed from the model catalog

deprecated
Qwen3.8 Max Preview

via alibaba-token-plan

What changed

  • statusbetadeprecated
context
Gemma 4 26B A4B IT

via kilo

What changed

  • Context262K131K−50%
  • Output limit16K33K2×
repriced
Nova 2 Lite (EU)

via amazon-bedrock

What changed

  • knowledge—2025-10
  • release date2025-12-012025-12-02
  • last updated2025-12-012025-12-02
  • Output limit64K66K1×
  • +1 more changes
repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.106/M$0.122/M▲ 15.1%
  • Output price$0.368/M$0.42/M▲ 14.1%
  • Cache read—$0.061/M
  • cache write$0.053/M—
context
Nova Lite

via edenai

What changed

  • Output limit8K10K1.2×
repriced
Kimi K2.5

via hyper

What changed

  • Cache read—$0.279/M
  • cache write$0.279/M—
new model
Mistral Large 3 675B Instruct 2512

via fireworks-ai

What changed

First observed in the model catalog

new model
GLM-5.2

via pioneer

What changed

First observed in the model catalog

repriced
Kimi K3

via openrouter

What changed

  • Input price$3/M$2.34/M▼ 22.0%
  • Output price$15/M$11.70/M▼ 22.0%
  • Cache read$0.3/M$0.261/M▼ 13.0%
new model
DeepSeek V4.1 Flash

via kilo

What changed

First observed in the model catalog

repriced
Nova 2 Lite (US)

via amazon-bedrock

What changed

  • knowledge—2025-10
  • release date2025-12-012025-12-02
  • last updated2025-12-012025-12-02
  • Output limit64K66K1×
  • +1 more changes
new model
Qwen Plus

via ofox

What changed

First observed in the model catalog

repriced
MoonshotAI Kimi Latest

via kilo

What changed

  • Input price$2.40/M$2.34/M▼ 2.5%
  • Output price$12/M$11.70/M▼ 2.5%
  • Cache read$0.24/M$0.261/M▲ 8.8%
new model
Gemma 4 31B IT (Consensus Protocol)

via llmgateway-providers

What changed

First observed in the model catalog

removed
DeepSeek V4 Flash Vision Exp (DeepSeek)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via opencode-go

What changed

  • Input price$0.22/M$0.15/M▼ 31.8%
  • Output price$0.66/M$0.6/M▼ 9.1%
  • Cache read$0.0070/M$0.0030/M▼ 57.1%
repriced
DeepSeek V4 Flash Vision Exp

via crossmodel

What changed

  • Input price$0.405/M$0.27/M▼ 33.3%
  • Output price$1.22/M$1.08/M▼ 11.1%
  • Cache read$0.013/M$0.0054/M▼ 60.0%
  • cache write$0.405/M$0.27/M▼ 33.3%
new model
Qwen Flash

via ofox

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via deepinfra

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via crossmodel

What changed

  • Input price$0.405/M$0.27/M▼ 33.3%
  • Output price$1.22/M$1.08/M▼ 11.1%
  • Cache read$0.013/M$0.0054/M▼ 60.0%
  • cache write$0.405/M$0.27/M▼ 33.3%
context
Nova Micro

via edenai

What changed

  • Output limit8K10K1.2×
deprecated
MiniMax-M2.7

via deepinfra

What changed

  • status—deprecated
new model
GLM-5.3

via bothub

What changed

First observed in the model catalog

new model
Qwen3.7 Plus

via ofox

What changed

First observed in the model catalog

repriced
Nova Micro

via amazon-bedrock

What changed

  • Output limit8K10K1.2×
  • cache write—$0.035/M
repriced
DeepSeek V4 Flash

via merge-gateway

What changed

  • Context1M1.05M1×
  • Input price$0.22/M$0.035/M▼ 84.1%
  • Output price$0.66/M$0.07/M▼ 89.4%
repriced
GLM-5

via hyper

What changed

  • Cache read—$0.43/M
  • cache write$0.43/M—
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context1.02M1.05M1×
  • Output limit128K944K7.4×
  • Input price$1.11/M$1/M▼ 10.2%
  • Output price$3.50/M$3.41/M▼ 2.5%
  • +1 more changes
new model
DeepSeek V4.1 Flash

via openrouter

What changed

First observed in the model catalog

context
GLM-5.2

via kilo

What changed

  • Context1.05M1M−5%
repriced
Mercury 2.5

via venice

What changed

  • Input price$0.25/M$0.05/M▼ 80.0%
  • Output price$0.938/M$0.187/M▼ 80.0%
  • Cache read$0.025/M$0.0050/M▼ 80.0%
new model
DeepSeek V4 Flash 0423

via merge-gateway

What changed

First observed in the model catalog

repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.458/M$0.404/M▼ 11.8%
  • Output price$1.71/M$1.50/M▼ 12.6%
  • Cache read—$0.202/M
  • cache write$0.229/M—
new model
Grok 4.20 (Reasoning)

via google-vertex

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct

via hyper

What changed

  • Cache read—$0.303/M
  • cache write$0.303/M—
new model
Qwen3.7 Max

via ofox

What changed

First observed in the model catalog

new model
Qwen3.8 27B

via ofox

What changed

First observed in the model catalog

repriced
GLM Latest

via openrouter

What changed

  • Output limit128K944K7.4×
  • Input price$1.11/M$1/M▼ 10.2%
  • Output price$3.50/M$3.41/M▼ 2.5%
  • Cache read$0.207/M$0.2/M▼ 3.2%

118 events
repriced
Kimi K2.7 Code

via llmgateway

What changed

  • Input price$0.89/M$0.95/M▲ 6.7%
  • Output price$3.71/M$4/M▲ 7.8%
  • Cache read$0.18/M$0.19/M▲ 5.6%
new model
Nova Lite (US)

via edenai

What changed

First observed in the model catalog

new model
Nova 2 Lite (EU)

via amazon-bedrock

What changed

First observed in the model catalog

deprecated
GPT-5.1 Chat

via zenmux

What changed

  • status—deprecated
repriced
GPT-5.2

via nearai

What changed

  • Input price$1.80/M$1.75/M▼ 2.8%
  • Output price$15.50/M$14/M▼ 9.7%
  • Cache read$0.18/M$0.175/M▼ 2.8%
new model
Qwen 3.8 Flash

via venice

What changed

First observed in the model catalog

new model
Nova Micro (US)

via amazon-bedrock

What changed

First observed in the model catalog

capability
Nex-N2.5-Mini (free)

via openrouter

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
repriced
GLM Flash Latest

via openrouter

What changed

  • Output limit944K131K−86%
  • Input price$0.075/M$0.07/M▼ 6.7%
  • Output price$0.25/M$0.233/M▼ 6.7%
  • Cache read$0.015/M$0.014/M▼ 6.7%
context
Kimi K2 Thinking

via kilo

What changed

  • Output limit236K100K−57%
new model
Gemma 4 12B StationKeeper

via nano-gpt

What changed

First observed in the model catalog

context
Inkling

via openrouter

What changed

  • Output limit472K33K−93%
new model
GPT-5.6 Terra (US)

via amazon-bedrock

What changed

First observed in the model catalog

new model
K2-Horizon-7B

via nano-gpt

What changed

First observed in the model catalog

repriced
GLM-5.2 Turbo (SCX.ai)

via llmgateway-providers

What changed

  • Input price$1.99/M$2.20/M▲ 10.6%
  • Output price$6.16/M$6.50/M▲ 5.5%
  • Cache read$0.4/M$0.45/M▲ 12.5%
new model
Nova Lite (EU)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GLM-5.3-Flash

via opencode-go

What changed

  • nameGLM-5.3-Flash (2x usage)GLM-5.3-Flash
  • Input price$0.075/M$0.15/M▲ 100.0%
  • Output price$0.25/M$0.5/M▲ 100.0%
  • Cache read$0.015/M$0.03/M▲ 100.0%
new model
Palmyra X5 (US)

via amazon-bedrock

What changed

First observed in the model catalog

context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.05M1.02M−2%
  • Output limit393K384K−2%
new model
Pixtral Large (25.02) (Amazon Bedrock)

via edenai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via deepinfra

What changed

  • Cache read$0.14/M$0.014/M▼ 90.0%
removed
Azure gpt-4-turbo

via nano-gpt

What changed

Removed from the model catalog

new model
Nova Micro (EU)

via amazon-bedrock

What changed

First observed in the model catalog

new model
Nova Lite (US)

via amazon-bedrock

What changed

First observed in the model catalog

removed
Gemma 4 31B IT

via nearai

What changed

Removed from the model catalog

new model
Gemma 4 12B Semancer

via nano-gpt

What changed

First observed in the model catalog

removed
Hermes 4 70B

via openrouter

What changed

Removed from the model catalog

new model
Nova Micro (US)

via edenai

What changed

First observed in the model catalog

repriced
Claude Sonnet 4.5 (latest)

via nearai

What changed

  • Output price$15.50/M$15/M▼ 3.2%
new model
Llama 3.1 70B

via merge-gateway

What changed

First observed in the model catalog

repriced
Qwen3.8 Max (SCX.ai)

via llmgateway-providers

What changed

  • Input price$1.81/M$2/M▲ 10.2%
  • Output price$5.45/M$6/M▲ 10.2%
  • Cache read$0.21/M$0.25/M▲ 19.0%
capability
Nova 2 Lite

via cortecs

What changed

  • namenova-2-liteNova 2 Lite
  • family—nova
  • release date2025-12-042025-12-01
  • last updated2025-12-042025-12-01
new model
Agnes 3.0 Flash

via nano-gpt

What changed

First observed in the model catalog

repriced
Kimi K3 (SCX.ai)

via llmgateway-providers

What changed

  • Input price$2.83/M$3.50/M▲ 23.7%
  • Output price$14.13/M$18/M▲ 27.4%
  • Cache read$0.28/M$0.35/M▲ 25.0%
context
Qwen3-VL 30B-A3B Instruct

via nearai

What changed

  • Context256K16K−94%
  • Output limit33K8K−75%
repriced
GPT OSS 120B (Scaleway)

via edenai

What changed

  • Input price$0.174/M$0.175/M▲ 0.3%
  • Output price$0.697/M$0.699/M▲ 0.3%
repriced
Qwen3 235B A22B Instruct 2507

via openrouter

What changed

  • Input price$0.09/M$0.22/M▲ 144.4%
  • Output price$0.55/M$0.88/M▲ 60.0%
context
Qwen3.5 122B-A10B

via kilo

What changed

  • Output limit82K66K−20%
capability
Nex AGI: Nex-N2.5-Mini (free)

via kilo

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
capability
GPT-6 Astra (US)

via amazon-bedrock

What changed

  • structured outputNoYesEnabled
new model
GPT-5.6 Luna (US)

via amazon-bedrock

What changed

First observed in the model catalog

removed
Azure o1

via nano-gpt

What changed

Removed from the model catalog

new model
GPT-5.6 Sol (US)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
Kimi K2.7 Code (SCX.ai)

via llmgateway-providers

What changed

  • Input price$0.89/M$0.95/M▲ 6.7%
  • Output price$3.71/M$4/M▲ 7.8%
  • Cache read$0.18/M$0.19/M▲ 5.6%
capability
Llama 3.1 70B

via wandb

What changed

  • knowledge—2023-12
repriced
DeepSeek Chat

via openrouter

What changed

  • Output limit16K16K−2%
  • Input price$0.32/M$0.257/M▼ 19.6%
  • Output price$0.89/M$1.03/M▲ 15.6%
new model
Nova Lite (CA)

via amazon-bedrock

What changed

First observed in the model catalog

context
Inkling

via kilo

What changed

  • Context524K1.05M2×
  • Output limit472K33K−93%
repriced
Z.ai: GLM Flash Latest

via kilo

What changed

  • Output limit944K131K−86%
  • Input price$0.075/M$0.07/M▼ 6.7%
  • Output price$0.25/M$0.233/M▼ 6.7%
  • Cache read$0.015/M$0.014/M▼ 6.7%
repriced
Kimi K2 Thinking

via openrouter

What changed

  • Output limit236K100K−57%
  • Cache read—$0.15/M
new model
Nova Lite

via edenai

What changed

First observed in the model catalog

repriced
Qwen3.5 122B-A10B

via openrouter

What changed

  • Output limit82K66K−20%
  • Input price$0.29/M$0.26/M▼ 10.3%
  • Output price$2.40/M$2.08/M▼ 13.3%
repriced
Qwen3.8 Max Preview

via llmgateway

What changed

  • Input price$1.81/M$2/M▲ 10.2%
  • Output price$5.45/M$6/M▲ 10.2%
  • Cache read$0.21/M$0.25/M▲ 19.0%
removed
Azure o3-mini

via nano-gpt

What changed

Removed from the model catalog

new model
Nova Pro (US)

via edenai

What changed

First observed in the model catalog

context
Qwen: Qwen3 30B A3B Instruct 2507

via kilo

What changed

  • Context128K262K2×
  • Output limit32K236K7.4×
new model
Pixtral Large (25.02) (US)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GLM-5.3-Flash

via openrouter

What changed

  • Input price$0.075/M$0.07/M▼ 6.7%
  • Output price$0.25/M$0.233/M▼ 6.7%
  • Cache read$0.015/M$0.014/M▼ 6.7%
repriced
Kimi K3

via llmgateway

What changed

  • Input price$2.83/M$3/M▲ 6.0%
  • Output price$14.13/M$15/M▲ 6.2%
  • Cache read$0.28/M$0.3/M▲ 7.1%
repriced
GLM-5.2 (SCX.ai)

via llmgateway-providers

What changed

  • Input price$0.55/M$0.8/M▲ 45.5%
  • Output price$1.78/M$2.55/M▲ 42.9%
  • Cache read$0.111/M$0.16/M▲ 44.1%
repriced
Qwen3 30B A3B Instruct 2507

via openrouter

What changed

  • Output limit32K236K7.4×
  • Input price$0.048/M$0.09/M▲ 86.9%
  • Output price$0.193/M$0.3/M▲ 55.4%
new model
Pixtral Large (25.02) (Amazon Bedrock, US)

via edenai

What changed

First observed in the model catalog

removed
Azure gpt-4o

via nano-gpt

What changed

Removed from the model catalog

new model
Nova Pro

via edenai

What changed

First observed in the model catalog

new model
Qwen 3.8 27B Queen

via nano-gpt

What changed

First observed in the model catalog

context
MiniMax-M2.5

via kilo

What changed

  • Context200K205K1×
  • Output limit128K131K1×
removed
Gemini 3 Pro Preview

via nearai

What changed

Removed from the model catalog

repriced
GLM-5.1

via hyper

What changed

  • Input price$1.36/M$1.39/M▲ 2.1%
  • Output price$4.27/M$4.36/M▲ 2.1%
  • cache write$0.679/M$0.693/M▲ 2.1%
new model
DeepSeek V4.1 Flash Thinking

via nano-gpt

What changed

First observed in the model catalog

new model
Ling 3.0 Flash VL

via nano-gpt

What changed

First observed in the model catalog

removed
GPT-OSS 120B

via nearai

What changed

Removed from the model catalog

repriced
Llama-3.3-70B-Instruct (IONOS)

via edenai

What changed

  • Input price$0.755/M$0.757/M▲ 0.3%
  • Output price$0.755/M$0.757/M▲ 0.3%
new model
Llama 3.1 70B Instruct (US)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GLM-5.3-Flash

via llmgateway

What changed

  • Input price$0.1/M$0.088/M▼ 12.0%
  • Cache read$0.02/M$0.025/M▲ 25.0%
repriced
GLM-5.2

via llmgateway

What changed

  • Input price$1.99/M$2.20/M▲ 10.6%
  • Output price$6.16/M$6.50/M▲ 5.5%
  • Cache read$0.4/M$0.45/M▲ 12.5%
new model
GLM-5.3

via greenpt

What changed

First observed in the model catalog

new model
Llama 3.3 70B Instruct (US)

via amazon-bedrock

What changed

First observed in the model catalog

capability
Llama-3.1-70B-Instruct

via openrouter

What changed

  • nameLlama 3.1 70B InstructLlama-3.1-70B-Instruct
  • knowledge2023-12-312023-12
context
Qwen 3.6 35B A3B FP8

via nearai

What changed

  • Output limit33K8K−75%
repriced
GLM-5.3 Flash (SCX.ai)

via llmgateway-providers

What changed

  • Input price$0.13/M$0.088/M▼ 32.3%
  • Output price$0.4/M$0.25/M▼ 37.5%
  • Cache read$0.024/M$0.025/M▲ 4.2%
new model
Nova 2 Lite (US)

via amazon-bedrock

What changed

First observed in the model catalog

new model
Nova Micro

via edenai

What changed

First observed in the model catalog

repriced
Mercury 2.5

via venice

What changed

  • Input price$0.05/M$0.25/M▲ 400.0%
  • Output price$0.187/M$0.938/M▲ 400.0%
  • Cache read$0.0050/M$0.025/M▲ 400.0%
repriced
GLM-5.2

via llmgateway

What changed

  • Input price$0.55/M$0.8/M▲ 45.5%
  • Output price$1.78/M$2.55/M▲ 42.9%
  • Cache read$0.111/M$0.16/M▲ 44.1%
new model
Nova Pro (EU)

via amazon-bedrock

What changed

First observed in the model catalog

new model
Nova Lite (APAC)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Output limit393K384K−2%
  • Input price$0.579/M$1.05/M▲ 81.1%
  • Output price$1.74/M$3.15/M▲ 81.1%
  • Cache read$0.058/M$0.035/M▼ 39.6%
removed
Qwen3.5 122B-A10B

via nearai

What changed

Removed from the model catalog

repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.178/M$0.18/M▲ 1.1%
  • Output price$0.68/M$0.61/M▼ 10.3%
  • cache write$0.089/M$0.09/M▲ 1.1%
capability
GPT-6 Astra (Global)

via amazon-bedrock

What changed

  • structured outputNoYesEnabled
new model
Nova Pro (APAC)

via amazon-bedrock

What changed

First observed in the model catalog

capability
FLUX.2 Klein 4B

via nearai

What changed

  • modalities.inputtext,imagetext
repriced
GLM-5.3-Flash

via crossmodel

What changed

  • Input price$0.075/M$0.15/M▲ 100.0%
  • Output price$0.25/M$0.5/M▲ 100.0%
  • Cache read$0.015/M$0.03/M▲ 100.0%
  • cache write$0.075/M$0.15/M▲ 100.0%
new model
Palmyra X4 (US)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
MiniMax-M2.5

via openrouter

What changed

  • Output limit128K131K1×
  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$1.08/M$1.20/M▲ 11.1%
  • Cache read$0.027/M$0.03/M▲ 11.1%
repriced
Whisper Large v3

via nearai

What changed

  • Output price$0/M$0.01/M
new model
Nova 2 Lite (Global)

via amazon-bedrock

What changed

First observed in the model catalog

capability
Pixtral Large (25.02)

via cortecs

What changed

  • namepixtral-large-2502Pixtral Large (25.02)
  • family—pixtral
  • release date2025-05-262025-04-08
  • last updated2025-05-262025-04-08
removed
Azure gpt-4o-mini

via nano-gpt

What changed

Removed from the model catalog

repriced
Kimi K3

via digitalocean

What changed

  • Input price$2.85/M$2.55/M▼ 10.5%
  • Output price$14.25/M$12.95/M▼ 9.1%
new model
Nova Micro (APAC)

via amazon-bedrock

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via greenpt

What changed

First observed in the model catalog

repriced
GLM-5.1 FP8

via nearai

What changed

  • Output limit131K16K−88%
  • Input price$0.85/M$1.40/M▲ 64.7%
  • Output price$3.30/M$4.40/M▲ 33.3%
repriced
Llama-3.3-70B-Instruct (Scaleway)

via edenai

What changed

  • Input price$1.05/M$1.05/M▲ 0.3%
  • Output price$1.05/M$1.05/M▲ 0.3%
new model
Claude Opus 4.7 (AU)

via amazon-bedrock

What changed

First observed in the model catalog

removed
Nous: Hermes 4 70B

via kilo

What changed

Removed from the model catalog

removed
Grok Imagine Video 1.5 Preview

via vercel

What changed

Removed from the model catalog

repriced
DeepSeek Chat

via kilo

What changed

  • Context164K128K−22%
  • Output limit16K16K−2%
  • Input price$0.32/M$0.257/M▼ 19.6%
  • Output price$0.89/M$1.03/M▲ 15.6%
repriced
Qwen3 Embedding 0.6B

via nearai

What changed

  • Context41K33K−20%
  • Output price$0/M$0.01/M
new model
Nova 2 Lite (JP)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GPT OSS 120B (IONOS)

via edenai

What changed

  • Input price$0.174/M$0.175/M▲ 0.3%
  • Output price$0.755/M$0.757/M▲ 0.3%
removed
Qwen3 30B-A3B Instruct 2507

via nearai

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731 (Scaleway)

via edenai

What changed

  • Input price$0.465/M$0.466/M▲ 0.3%
  • Output price$0.929/M$0.932/M▲ 0.3%
new model
Llama 3.1 8B Instruct (US)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GLM-5.3 (SCX.ai)

via llmgateway-providers

What changed

  • Input price$1.30/M$1.40/M▲ 7.7%
  • Output price$4/M$4.40/M▲ 10.0%
  • Cache read$0.25/M$0.26/M▲ 4.0%
new model
Pixtral Large (25.02) (EU)

via amazon-bedrock

What changed

First observed in the model catalog

capability
Llama-3.1-70B-Instruct

via kilo

What changed

  • nameMeta: Llama 3.1 70B InstructLlama-3.1-70B-Instruct
  • open weightsNoYesEnabled
  • knowledge—2023-12
new model
Nova Pro (US)

via amazon-bedrock

What changed

First observed in the model catalog

203 events
capability
Nemotron 3 Super 120B A12B

via kilo

What changed

  • structured outputYesNoRemoved
capability
MiniMax-M2.7

via 302ai

What changed

  • descriptionMiniMax model for chat, coding, office work, and agentic tasksOpen MiniMax flagship for coding agents, office automation, and complex environments
  • family—minimax
  • reasoningNoYesEnabled
  • open weightsNoYesEnabled
  • +2 more changes
capability
GLM 4.6V

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
new model
gemini-3.5-flash-thinking

via 302ai

What changed

First observed in the model catalog

new model
Kimi K2.5

via 302ai

What changed

First observed in the model catalog

removed
Deepseek-Chat

via 302ai

What changed

Removed from the model catalog

removed
gpt-5.4-nano-2026-03-17

via 302ai

What changed

Removed from the model catalog

new model
Claude Fable 5

via 302ai

What changed

First observed in the model catalog

removed
Yi Large

via nano-gpt

What changed

Removed from the model catalog

repriced
thinkingcap-qwen3.6-27b@eu

via requesty

What changed

  • Output price$3/M$2.60/M▼ 13.3%
  • Cache read$0.26/M$0.05/M▼ 80.8%
new model
GPT-5.6 Luna

via 302ai

What changed

First observed in the model catalog

removed
claude-opus-4-6-thinking

via 302ai

What changed

Removed from the model catalog

capability
Gemini 2.5 Pro

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
new model
Gemini 3.1 Flash Lite

via 302ai

What changed

First observed in the model catalog

new model
Qwen3.6 35B-A3B

via 302ai

What changed

First observed in the model catalog

new model
GLM-5.3

via privatemode-ai

What changed

First observed in the model catalog

new model
Mercury 2.5

via vercel

What changed

First observed in the model catalog

capability
Qwen3.5 9B

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
repriced
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.02M1.05M1×
  • Output limit384K393K1×
  • Cache read$0.132/M$0.044/M▼ 66.7%
new model
GPT-6 Astra (Global)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
GPT OSS 120B (Scaleway)

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.1%
  • Output price$0.697/M$0.697/M▼ 0.1%
repriced
Qwen3-Next 80B-A3B Instruct

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.1/M$0.09/M▼ 10.0%
  • Cache read$0.07/M—
repriced
Qwen3.8 27B (Consensus Protocol)

via llmgateway-providers

What changed

  • Input price$0.41/M$0.2/M▼ 51.2%
  • Output price$2.50/M$2/M▼ 20.0%
  • Cache read$0.08/M$0.05/M▼ 37.5%
new model
Qwen3.8 Flash

via 302ai

What changed

First observed in the model catalog

new model
claude-opus-4-7-thinking

via 302ai

What changed

First observed in the model catalog

capability
Gemini 3.1 Pro (Preview High)

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
repriced
DeepSeek V4 Flash 0731 (FlexAI)

via edenai

What changed

  • Input price$0.03/M$0.065/M▲ 116.7%
  • Output price$0.1/M$0.18/M▲ 80.0%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Output limit131K944K7.2×
  • Input price$0.14/M$0.065/M▼ 53.6%
  • Output price$0.28/M$0.18/M▼ 35.7%
  • Cache read$0.028/M$0.016/M▼ 42.9%
new model
Qwen3.6 Plus

via 302ai

What changed

First observed in the model catalog

repriced
GPT OSS 20B (FlexAI)

via edenai

What changed

  • Input price$0.02/M$0.03/M▲ 50.0%
  • Output price$0.1/M$0.13/M▲ 30.0%
repriced
MoonshotAI Kimi Latest

via kilo

What changed

  • Input price$2.50/M$2.40/M▼ 4.0%
  • Output price$14/M$12/M▼ 14.3%
  • Cache read$0.29/M$0.24/M▼ 17.2%
capability
GLM 5V Turbo Thinking

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
repriced
GLM Latest

via openrouter

What changed

  • Output limit944K128K−86%
  • Input price$1.12/M$1.11/M▼ 0.6%
  • Output price$3.52/M$3.50/M▼ 0.6%
  • Cache read$0.208/M$0.207/M▼ 0.6%
context
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Output limit384K944K2.5×
new model
Kimi K3

via 302ai

What changed

First observed in the model catalog

new model
gpt-5.6-sol-pro

via 302ai

What changed

First observed in the model catalog

context
Qwen3.8 2.4T A95B

via kilo

What changed

  • Output limit262K131K−50%
new model
Claude Opus 5

via 302ai

What changed

First observed in the model catalog

new model
Nex AGI: Nex-N2.5-Pro (free)

via kilo

What changed

First observed in the model catalog

new model
Gemini 3.1 Flash Lite Preview

via 302ai

What changed

First observed in the model catalog

new model
GPT-5.6 Terra

via 302ai

What changed

First observed in the model catalog

new model
Grok 4.5

via 302ai

What changed

First observed in the model catalog

removed
MiniMax-M2.7-highspeed

via 302ai

What changed

Removed from the model catalog

capability
Muse Spark 1.1

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video,audio
removed
gpt-5.4-mini-2026-03-17

via 302ai

What changed

Removed from the model catalog

context
Qwen3.5 397B-A17B

via kilo

What changed

  • Output limit66K236K3.6×
new model
Qwen3.5 35B-A3B

via 302ai

What changed

First observed in the model catalog

removed
claude-3-5-haiku-latest

via 302ai

What changed

Removed from the model catalog

capability
Gemini 3.1 Pro (Preview)

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
new model
GPT-5.6 Sol

via 302ai

What changed

First observed in the model catalog

new model
Qwen3.5 Plus

via 302ai

What changed

First observed in the model catalog

capability
Gemini 2.5 Flash Lite

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
new model
qwen3.7-max-2026-06-08

via 302ai

What changed

First observed in the model catalog

new model
DeepSeek V4.1 Flash

via nano-gpt

What changed

First observed in the model catalog

capability
Qwen3.8 27B Thinking

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
removed
claude-opus-4-5

via 302ai

What changed

Removed from the model catalog

removed
Nex N2 Mini

via nano-gpt

What changed

Removed from the model catalog

capability
Inkling Small Thinking

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,audio
removed
GLM 5.3 (50% off)

via vercel

What changed

Removed from the model catalog

removed
kimi-k2-thinking-turbo

via 302ai

What changed

Removed from the model catalog

repriced
Llama-3.3-70B-Instruct (Scaleway)

via edenai

What changed

  • Input price$1.05/M$1.05/M▼ 0.1%
  • Output price$1.05/M$1.05/M▼ 0.1%
removed
grok-4.20-beta-0309-non-reasoning

via 302ai

What changed

Removed from the model catalog

repriced
Qwen3.8 27B

via llmgateway

What changed

  • Input price$0.41/M$0.2/M▼ 51.2%
  • Output price$2.50/M$2/M▼ 20.0%
  • Cache read$0.08/M$0.05/M▼ 37.5%
removed
glm-4.5-airx

via 302ai

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Output limit384K944K2.5×
  • Input price$0.44/M$0.22/M▼ 50.0%
  • Output price$1.32/M$0.66/M▼ 50.0%
  • Cache read$0.014/M$0.0070/M▼ 50.0%
capability
Gemini 3 Flash Thinking

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
new model
Mercury 2.5

via venice

What changed

First observed in the model catalog

removed
Inception: Mercury 2.5 Preview

via kilo

What changed

Removed from the model catalog

removed
claude-sonnet-4-20250514

via 302ai

What changed

Removed from the model catalog

capability
ZDev

via zeldoc

What changed

  • knowledge2025-01—
  • release date2026-04-152026-08-26
  • last updated2026-04-152026-08-26
repriced
GLM 5.3

via vercel

What changed

  • Input price$0.7/M$1.40/M▲ 100.0%
  • Output price$2.20/M$4.40/M▲ 100.0%
  • Cache read$0.13/M$0.14/M▲ 7.7%
new model
GLM-5.3

via 302ai

What changed

First observed in the model catalog

capability
Gemma 4 31B Thinking

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
repriced
GLM-5.1

via 302ai

What changed

  • nameglm-5.1GLM-5.1
  • descriptionFlagship GLM model for hybrid reasoning, coding, and agentic engineeringStrong GLM coding model for agentic engineering, terminals, and repository generation
  • open weightsNoYesEnabled
  • release date2026-04-102026-04-07
  • +3 more changes
repriced
Gemini 3.8 Flash

via nano-gpt

What changed

  • cache write$0.075/M$0.042/M▼ 44.4%
new model
GPT Image 2.5 Flare

via vercel

What changed

First observed in the model catalog

capability
Gemma 4 26B A4B Thinking

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
new model
Claude Fable 5.1

via 302ai

What changed

First observed in the model catalog

context
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit262K131K−50%
repriced
Kimi K2.6

via llmgateway

What changed

  • Input price$0.22/M$0.6/M▲ 172.7%
  • Output price$1.14/M$3.05/M▲ 168.2%
  • Cache read$0.048/M$0.13/M▲ 170.8%
repriced
DeepSeek V4 Flash 0731 (EU)

via requesty

What changed

  • Input price$0.14/M$0.28/M▲ 100.0%
  • Output price$0.28/M$0.56/M▲ 100.0%
new model
claude-opus-5-thinking

via 302ai

What changed

First observed in the model catalog

new model
GLM (latest)

via privatemode-ai

What changed

First observed in the model catalog

new model
Grok 4.3

via 302ai

What changed

First observed in the model catalog

removed
Nvidia Nemotron 3 Nano Omni

via nano-gpt

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731 (Scaleway)

via edenai

What changed

  • Input price$0.465/M$0.465/M▼ 0.1%
  • Output price$0.93/M$0.929/M▼ 0.1%
capability
Qwen3.8 27B

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
removed
Nex-N2-Pro

via openrouter

What changed

Removed from the model catalog

new model
gpt-5.6-terra-pro

via 302ai

What changed

First observed in the model catalog

new model
Gemini 3.7 Flash

via 302ai

What changed

First observed in the model catalog

removed
claude-sonnet-4-5

via 302ai

What changed

Removed from the model catalog

new model
DeepSeek V4.1 Flash Beta

via vercel

What changed

First observed in the model catalog

removed
Nex AGI: Nex-N2-Mini (retires Sep 8)

via kilo

What changed

Removed from the model catalog

context
DeepSeek V4 Flash 0731

via kilo

What changed

  • Output limit131K944K7.2×
capability
Gemini 3.1 Pro (Preview Custom Tools)

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
new model
Gemini 3.8 Flash

via cortecs

What changed

First observed in the model catalog

new model
GPT-6 Astra

via requesty

What changed

First observed in the model catalog

capability
Gemini 3.5 Flash

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.188/M$0.178/M▼ 5.3%
  • Output price$0.7/M$0.68/M▼ 2.9%
  • cache write$0.094/M$0.089/M▼ 5.3%
new model
gpt-5.6-luna-pro

via 302ai

What changed

First observed in the model catalog

new model
GPT-5.5

via 302ai

What changed

First observed in the model catalog

removed
Qwen3.8 27B (free)

via orcarouter

What changed

Removed from the model catalog

removed
glm-for-coding

via 302ai

What changed

Removed from the model catalog

repriced
Kimi K3 (EU)

via requesty

What changed

  • Input price$2.25/M$3/M▲ 33.3%
  • Output price$11.25/M$15/M▲ 33.3%
  • Cache read$0.225/M$0.45/M▲ 100.0%
context
Qwen3-Next 80B-A3B Instruct

via kilo

What changed

  • Output limit236K16K−93%
repriced
GLM-5.3-Flash

via requesty

What changed

  • Output price$0.5/M$0.6/M▲ 20.0%
new model
Gemini 3.5 Flash Lite

via 302ai

What changed

First observed in the model catalog

repriced
GLM-5.3-Flash (EU)

via requesty

What changed

  • Output price$0.5/M$0.6/M▲ 20.0%
new model
Nex AGI: Nex-N2.5-Mini (free)

via kilo

What changed

First observed in the model catalog

new model
GPT-6 Astra

via 302ai

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.11/M$0.106/M▼ 3.6%
  • Output price$0.408/M$0.368/M▼ 9.8%
  • cache write$0.055/M$0.053/M▼ 3.6%
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context1.05M1.02M−2%
  • Output limit944K128K−86%
  • Input price$1.12/M$1.11/M▼ 0.6%
  • Output price$3.52/M$3.50/M▼ 0.6%
  • +1 more changes
removed
glm-4.5-air

via 302ai

What changed

Removed from the model catalog

new model
Kimi K2.6

via 302ai

What changed

First observed in the model catalog

new model
Kimi K2.7 Code

via 302ai

What changed

First observed in the model catalog

context
MiniMax-M2.1

via 302ai

What changed

  • descriptionMiniMax model for chat, coding, office work, and agentic tasksEarlier MiniMax agent model for practical coding and productivity tasks
  • family—minimax
  • reasoningNoYesEnabled
  • open weightsNoYesEnabled
  • +3 more changes
capability
Gemma 4 26B A4B

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
repriced
GLM-4.6

via openrouter

What changed

  • Output limit131K16K−88%
  • Input price$0.55/M$0.43/M▼ 21.8%
  • Output price$2.20/M$1.75/M▼ 20.5%
  • Cache read$0.11/M$0.08/M▼ 27.3%
removed
glm-4.5-x

via 302ai

What changed

Removed from the model catalog

repriced
GLM-5.1

via hyper

What changed

  • Input price$1.33/M$1.36/M▲ 2.0%
  • Output price$4.31/M$4.27/M▼ 1.0%
  • cache write$0.666/M$0.679/M▲ 2.0%
capability
Gemini 2.5 Flash (No Thinking)

via nano-gpt

What changed

  • modalities.inputtext,image,audio,pdftext,image,video,audio,pdf
new model
GPT-6 Astra (US)

via amazon-bedrock

What changed

First observed in the model catalog

removed
Laguna XS 2.1

via nano-gpt

What changed

Removed from the model catalog

repriced
MiniMax M1

via openrouter

What changed

  • Input price$0.4/M$0.55/M▲ 37.5%
removed
claude-opus-4-20250514

via 302ai

What changed

Removed from the model catalog

repriced
Gemini 3.7 Flash

via nano-gpt

What changed

  • cache write$0.075/M$0.042/M▼ 44.4%
new model
Mercury 2.5

via openrouter

What changed

First observed in the model catalog

new model
Claude Opus 4.8

via 302ai

What changed

First observed in the model catalog

repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.47/M$0.458/M▼ 2.6%
  • Output price$1.76/M$1.71/M▼ 2.7%
  • cache write$0.235/M$0.229/M▼ 2.6%
repriced
DeepSeek V4 Flash 0731

via requesty

What changed

  • Input price$0.14/M$0.28/M▲ 100.0%
  • Output price$0.28/M$0.56/M▲ 100.0%
context
Nemotron 3 Super 120B A12B

via openrouter

What changed

  • structured outputYesNoRemoved
  • Context1M262K−74%
capability
Inkling Small

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,audio
capability
Kimi K2 0711 Fast

via nano-gpt

What changed

  • tool callYesNoRemoved
repriced
Granite 4.2 8B

via openrouter

What changed

  • Input price$0.1/M$0.06/M▼ 40.0%
  • Output price$0.15/M$0.25/M▲ 66.7%
  • Cache read$0.05/M$0.015/M▼ 70.0%
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Output limit384K393K1×
  • Input price$1.05/M$0.579/M▼ 44.8%
  • Output price$3.15/M$1.74/M▼ 44.8%
  • Cache read$0.035/M$0.058/M▲ 65.7%
removed
Qwen-Flash

via 302ai

What changed

Removed from the model catalog

repriced
GPT OSS 120B (IONOS)

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.1%
  • Output price$0.755/M$0.755/M▼ 0.1%
repriced
Sarvam 105B

via nano-gpt

What changed

  • Input price$0.045/M$0.054/M▲ 20.0%
  • Output price$0.177/M$0.212/M▲ 20.0%
  • Cache read$0.028/M$0.034/M▲ 20.0%
removed
glm-4.7-flashx

via 302ai

What changed

Removed from the model catalog

removed
Kimi K2.5

via crossmodel

What changed

Removed from the model catalog

new model
GPT Image 2.5 Sunburst

via vercel

What changed

First observed in the model catalog

removed
chatgpt-4o-latest

via 302ai

What changed

Removed from the model catalog

repriced
GPT OSS 120B (FlexAI)

via edenai

What changed

  • Input price$0.039/M$0.037/M▼ 5.1%
  • Output price$0.1/M$0.17/M▲ 70.0%
removed
Kimi K2.6 (Gonka24)

via llmgateway-providers

What changed

Removed from the model catalog

new model
Nex-N2.5-Pro (free)

via openrouter

What changed

First observed in the model catalog

capability
Kimi K2 0711 Instruct FP4

via nano-gpt

What changed

  • tool callYesNoRemoved
capability
MiMo V2.5 Thinking

via nano-gpt

What changed

  • modalities.inputtext,image,videotext,image,audio,video
new model
GLM-5.3-Flash

via 302ai

What changed

First observed in the model catalog

removed
Nex-N2-Mini

via openrouter

What changed

Removed from the model catalog

removed
claude-opus-4-5-20251101-thinking

via 302ai

What changed

Removed from the model catalog

capability
MiMo V2.5

via nano-gpt

What changed

  • modalities.inputtext,image,videotext,image,audio,video
new model
Gemini 3.1 Pro Preview

via 302ai

What changed

First observed in the model catalog

repriced
MoonshotAI Kimi Latest

via openrouter

What changed

  • Input price$2.50/M$2.40/M▼ 4.0%
  • Output price$14/M$12/M▼ 14.3%
  • Cache read$0.29/M$0.24/M▼ 17.2%
new model
Qwen3.7 Plus

via 302ai

What changed

First observed in the model catalog

repriced
Z.ai: GLM Flash Latest

via kilo

What changed

  • Output limit131K944K7.2×
  • Input price$0.071/M$0.075/M▲ 5.3%
  • Output price$0.237/M$0.25/M▲ 5.3%
  • Cache read$0.014/M$0.015/M▲ 5.3%
capability
Gemini 2.5 Flash

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
removed
claude-3-5-haiku-20241022

via 302ai

What changed

Removed from the model catalog

new model
o3

via 302ai

What changed

First observed in the model catalog

capability
GPT OSS 20B (Consensus Protocol)

via llmgateway-providers

What changed

  • attachmentNoYesEnabled
removed
Yi Medium 200k

via nano-gpt

What changed

Removed from the model catalog

new model
OpenAI GPT-6 Astra

via digitalocean

What changed

First observed in the model catalog

removed
Qwen-Max-Latest

via 302ai

What changed

Removed from the model catalog

new model
Qwen3.7 Max

via 302ai

What changed

First observed in the model catalog

removed
Deepseek-Reasoner

via 302ai

What changed

Removed from the model catalog

repriced
thinkingcap-qwen3.6-27b

via requesty

What changed

  • Output price$3/M$2.60/M▼ 13.3%
  • Cache read$0.26/M$0.05/M▼ 80.8%
removed
Mercury 2.5 Preview

via openrouter

What changed

Removed from the model catalog

removed
Nex AGI: Nex-N2-Pro (retires Sep 8)

via kilo

What changed

Removed from the model catalog

new model
Gemini 3.8 Flash

via 302ai

What changed

First observed in the model catalog

capability
Gemma 4 31B

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
new model
GLM-5.2

via 302ai

What changed

First observed in the model catalog

new model
MiniMax-M3

via 302ai

What changed

First observed in the model catalog

repriced
GLM Flash Latest

via openrouter

What changed

  • Output limit131K944K7.2×
  • Input price$0.071/M$0.075/M▲ 5.3%
  • Output price$0.237/M$0.25/M▲ 5.3%
  • Cache read$0.014/M$0.015/M▲ 5.3%
capability
GLM 5V Turbo

via nano-gpt

What changed

  • modalities.inputtext,imagetext,image,video
capability
Gemini 3.5 Flash Thinking

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
new model
Qwen3.8 Max

via 302ai

What changed

First observed in the model catalog

repriced
GLM-4.6

via kilo

What changed

  • Context205K198K−3%
  • Output limit131K16K−88%
  • Input price$0.55/M$0.43/M▼ 21.8%
  • Output price$2.20/M$1.75/M▼ 20.5%
  • +1 more changes
repriced
GLM 5.3 Flash Uncensored

via nano-gpt

What changed

  • Input price$0.07/M$0.35/M▲ 400.0%
  • Output price$0.21/M$1.40/M▲ 566.7%
  • Cache read$0.014/M$0.175/M▲ 1150.0%
capability
Gemini 3 Flash (Preview)

via nano-gpt

What changed

  • modalities.inputtext,image,audiotext,image,video,audio
removed
gpt-5.4-pro

via 302ai

What changed

Removed from the model catalog

new model
GLM-5.3-Flash (free)

via orcarouter

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct (IONOS)

via edenai

What changed

  • Input price$0.755/M$0.755/M▼ 0.1%
  • Output price$0.755/M$0.755/M▼ 0.1%
removed
grok-4.20-multi-agent-beta-0309

via 302ai

What changed

Removed from the model catalog

new model
Qwen3.6 Flash

via 302ai

What changed

First observed in the model catalog

new model
Gemini 3.5 Flash

via 302ai

What changed

First observed in the model catalog

new model
GPT-5.3 Chat (latest)

via 302ai

What changed

First observed in the model catalog

repriced
Qwen3.5 397B-A17B

via openrouter

What changed

  • Output limit66K236K3.6×
  • Input price$0.39/M$0.55/M▲ 41.0%
  • Output price$2.34/M$3.50/M▲ 49.6%
  • Cache read—$0.225/M
new model
Qwen3 8B TEE

via nano-gpt

What changed

First observed in the model catalog

new model
Gemini 3.6 Flash

via 302ai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via requesty

What changed

  • Input price$0.14/M$0.28/M▲ 100.0%
  • Output price$0.28/M$0.56/M▲ 100.0%
removed
Qwen-Plus

via 302ai

What changed

Removed from the model catalog

new model
Grok 4.6

via 302ai

What changed

First observed in the model catalog

removed
MiMo V2.5 Pro UltraSpeed

via vercel

What changed

Removed from the model catalog

new model
Nex-N2.5-Mini (free)

via openrouter

What changed

First observed in the model catalog

removed
claude-opus-4-6

via 302ai

What changed

Removed from the model catalog

new model
GPT-6 Astra

via amazon-bedrock

What changed

First observed in the model catalog

removed
doubao-seed-code-preview-251028

via 302ai

What changed

Removed from the model catalog

new model
Inception: Mercury 2.5

via kilo

What changed

First observed in the model catalog

repriced
Kimi K3

via requesty

What changed

  • Input price$2.25/M$3/M▲ 33.3%
  • Output price$11.25/M$15/M▲ 33.3%
  • Cache read$0.225/M$0.45/M▲ 100.0%
removed
grok-4-1-fast-non-reasoning

via 302ai

What changed

Removed from the model catalog

removed
Nex N2 Pro

via nano-gpt

What changed

Removed from the model catalog

new model
Claude Sonnet 5

via 302ai

What changed

First observed in the model catalog

new model
MiniMax-M2.5

via 302ai

What changed

First observed in the model catalog

capability
ByteDance Seed 2.0 Lite

via nano-gpt

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image,video

113 events
capability
GLM 5.3-Flash

via crof

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via aihubmix

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash (EU)

via requesty

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via zhipuai

What changed

  • open weightsNoYesEnabled
repriced
MoonshotAI Kimi Latest

via openrouter

What changed

  • Input price$2.55/M$2.50/M▼ 2.0%
  • Output price$12.75/M$14/M▲ 9.8%
  • Cache read$0.256/M$0.29/M▲ 13.3%
removed
Qwen3-Next 80B-A3B (Thinking)

via cortecs

What changed

Removed from the model catalog

repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.484/M$0.47/M▼ 2.9%
  • Output price$1.85/M$1.76/M▼ 5.0%
  • cache write$0.242/M$0.235/M▼ 2.9%
repriced
MoonshotAI Kimi Latest

via kilo

What changed

  • Input price$2.55/M$2.50/M▼ 2.0%
  • Output price$12.75/M$14/M▲ 9.8%
  • Cache read$0.256/M$0.29/M▲ 13.3%
capability
GLM 5.3 Flash

via nano-gpt

What changed

  • open weightsNoYesEnabled
repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.106/M$0.11/M▲ 3.8%
  • Output price$0.368/M$0.408/M▲ 10.9%
  • cache write$0.053/M$0.055/M▲ 3.8%
capability
GLM-5.3-Flash

via hyper

What changed

  • open weightsNoYesEnabled
capability
GLM 5.3 Flash

via vercel

What changed

  • open weightsNoYesEnabled
context
Qwen: Qwen3 14B

via kilo

What changed

  • Context41K131K3.2×
  • Output limit16K8K−50%
removed
MiniMax M2.7 (free)

via openrouter

What changed

Removed from the model catalog

capability
GLM-5.3-Flash

via zhipuai-coding-plan

What changed

  • open weightsNoYesEnabled
repriced
GLM-4.7-Flash

via kilo

What changed

  • Context203K131K−35%
  • Output limit16K118K7.2×
  • Input price$0.06/M$0.06/M▲ 0.8%
  • Cache read$0.01/M—
repriced
GLM Latest

via openrouter

What changed

  • Output limit236K944K4×
  • Input price$1.17/M$1.12/M▼ 4.3%
  • Output price$3.96/M$3.52/M▼ 11.1%
  • Cache read$0.234/M$0.208/M▼ 11.1%
capability
GLM-5.3 Flash (SCX.ai)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3 Flash (NovitaAI)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek V3.1

via openrouter

What changed

  • Output limit145K33K−77%
  • Input price$0.55/M$0.25/M▼ 54.5%
  • Output price$1.65/M$0.95/M▼ 42.4%
  • Cache read$0.55/M$0.13/M▼ 76.4%
repriced
GLM-5.3

via cortecs

What changed

  • Input price$1.40/M$1.11/M▼ 20.4%
  • Output price$4.40/M$3.90/M▼ 11.4%
  • Cache read$0.26/M$0.279/M▲ 7.3%
capability
GLM-5.3-Flash

via synthetic

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$1.04/M$0.955/M▼ 7.7%
  • Output price$2.07/M$1.91/M▼ 7.7%
  • Cache read$0.086/M$0.08/M▼ 7.7%
repriced
GLM Flash Latest

via openrouter

What changed

  • Output limit944K131K−86%
  • Input price$0.075/M$0.071/M▼ 5.0%
  • Output price$0.25/M$0.237/M▼ 5.0%
  • Cache read$0.015/M$0.014/M▼ 5.0%
capability
GLM-5.3 Flash (Consensus Protocol)

via llmgateway-providers

What changed

  • attachmentNoYesEnabled
  • open weightsNoYesEnabled
  • modalities.inputtexttext,image,video,pdf
repriced
Z.ai: GLM Flash Latest

via kilo

What changed

  • Output limit944K131K−86%
  • Input price$0.075/M$0.071/M▼ 5.0%
  • Output price$0.25/M$0.237/M▼ 5.0%
  • Cache read$0.015/M$0.014/M▼ 5.0%
new model
Laguna XS 2.1

via nano-gpt

What changed

First observed in the model catalog

capability
GLM 5.3 Flash

via modal

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek V3 0324

via openrouter

What changed

  • Input price$0.25/M$0.29/M▲ 16.0%
  • Output price$1/M$1.14/M▲ 14.0%
  • Cache read—$0.11/M
capability
GLM-5.3-Flash (2x usage)

via opencode-go

What changed

  • open weightsNoYesEnabled
repriced
Kimi K2.7 Code

via openrouter

What changed

  • Input price$0.66/M$0.71/M▲ 7.6%
  • Output price$3.40/M$3.50/M▲ 2.9%
  • Cache read$0.18/M$0.15/M▼ 16.7%
repriced
GLM 5.2 Thinking TEE

via nano-gpt

What changed

  • Output price$4.60/M$4.40/M▼ 4.3%
  • Cache read$0.5/M$0.7/M▲ 40.0%
context
DeepSeek: DeepSeek V3.1

via kilo

What changed

  • Context161K164K1×
  • Output limit145K33K−77%
capability
GLM-5.3-Flash

via nebius

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek: DeepSeek V3 0324

via kilo

What changed

  • Input price$0.25/M$0.29/M▲ 16.0%
  • Output price$1/M$1.14/M▲ 14.0%
  • Cache read—$0.11/M
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.26/M$1.33/M▲ 5.5%
  • Output price$4.13/M$4.31/M▲ 4.4%
  • cache write$0.631/M$0.666/M▲ 5.5%
removed
llama-3.1-nemotron-ultra-253b-v1

via cortecs

What changed

Removed from the model catalog

new model
GPT-6 Astra

via vivgrid

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via tokengo

What changed

  • open weightsNoYesEnabled
context
Kimi K2 Thinking

via kilo

What changed

  • Output limit100K236K2.4×
capability
GLM-5.3-Flash

via cortecs

What changed

  • open weightsNoYesEnabled
context
GLM 5.3

via fireworks-ai

What changed

  • last updated2026-09-042026-09-07
  • Context1M1.05M1×
  • Output limit131K262K2×
capability
GLM-5.3-Flash

via requesty

What changed

  • open weightsNoYesEnabled
capability
GLM 5.3 Flash

via baseten

What changed

  • open weightsNoYesEnabled
removed
MiniMax M3 (Free)

via vercel

What changed

Removed from the model catalog

repriced
OpenAI: GPT-5.6 Sol (50% off)

via kilo

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
removed
Qwen3 8B TEE

via nano-gpt

What changed

Removed from the model catalog

new model
Gemini 3.8 Flash

via vivgrid

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via kenari

What changed

  • open weightsNoYesEnabled
removed
cosmos3-super-reasoner

via cortecs

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit944K393K−58%
  • Input price$0.045/M$0.05/M▲ 11.1%
  • Output price$0.09/M$0.16/M▲ 77.8%
  • Cache read$0.0090/M$0.013/M▲ 44.4%
capability
GLM-5.3 Flash (Z AI)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via orcarouter

What changed

  • open weightsNoYesEnabled
repriced
GLM 5.3 Flash Uncensored

via nano-gpt

What changed

  • Input price$0.35/M$0.07/M▼ 80.0%
  • Output price$1.40/M$0.21/M▼ 85.0%
  • Cache read$0.175/M$0.014/M▼ 92.0%
repriced
GLM-4.6

via openrouter

What changed

  • Output limit16K131K8×
  • Input price$0.43/M$0.55/M▲ 27.9%
  • Output price$1.75/M$2.20/M▲ 25.7%
  • Cache read$0.08/M$0.11/M▲ 37.5%
context
GLM 5.3 Flash

via fireworks-ai

What changed

  • open weightsNoYesEnabled
  • last updated2026-09-042026-09-07
  • Context1M1.05M1×
repriced
GLM-4.6

via kilo

What changed

  • Context198K205K1×
  • Output limit16K131K8×
  • Input price$0.43/M$0.55/M▲ 27.9%
  • Output price$1.75/M$2.20/M▲ 25.7%
  • +1 more changes
capability
GLM-5.3-Flash

via vancine

What changed

  • open weightsNoYesEnabled
deprecated
Kimi K2.5

via deepinfra

What changed

  • status—deprecated
capability
GLM 5.3 Flash

via above

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via opencode

What changed

  • open weightsNoYesEnabled
removed
nvidia-nemotron-3-nano-omni

via cortecs

What changed

Removed from the model catalog

new model
DeepSeek V4 Flash Vision Exp Uncensored

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via deepinfra

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via zai

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$1.12/M$1.05/M▼ 6.4%
  • Output price$3.36/M$3.15/M▼ 6.4%
  • Cache read$0.037/M$0.035/M▼ 6.4%
repriced
Qwen3 14B

via openrouter

What changed

  • Output limit16K8K−50%
  • Input price$0.12/M$0.228/M▲ 89.6%
  • Output price$0.24/M$0.91/M▲ 279.2%
removed
MiniMax: MiniMax M2.7 (free)

via kilo

What changed

Removed from the model catalog

capability
GLM-5.3 Flash

via merge-gateway

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via runinfra

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.09/M$0.089/M▼ 1.4%
  • Output price$0.18/M$0.177/M▼ 1.4%
  • Cache read$0.018/M$0.018/M▼ 1.4%
capability
Qwen3 30B A3B

via kilo

What changed

  • structured outputNoYesEnabled
capability
GLM-5.3-Flash

via openrouter

What changed

  • open weightsNoYesEnabled
new model
Grok 4.6 (Global)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit944K393K−58%
  • Input price$0.045/M$0.05/M▲ 11.1%
  • Output price$0.09/M$0.16/M▲ 77.8%
  • Cache read$0.0090/M$0.013/M▲ 44.4%
capability
GLM-5.3-Flash

via tinfoil

What changed

  • open weightsNoYesEnabled
capability
GLM5.3 Flash

via digitalocean

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via edenai

What changed

  • open weightsNoYesEnabled
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context262K1.05M4×
  • Output limit236K944K4×
  • Input price$1.17/M$1.12/M▼ 4.3%
  • Output price$3.96/M$3.52/M▼ 11.1%
  • +1 more changes
new model
DeepSeek V4 Flash 0731 (FlexAI)

via edenai

What changed

First observed in the model catalog

repriced
Magistral Medium (latest)

via edenai

What changed

  • Input price$2/M$1.50/M▼ 25.0%
  • Output price$5/M$7.50/M▲ 50.0%
  • Cache read—$0.15/M
capability
GLM-5.3-Flash

via crossmodel

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via berget

What changed

  • open weightsNoYesEnabled
new model
DeepSeek V4 Pro 0813

via vivgrid

What changed

First observed in the model catalog

repriced
Kimi K2 Thinking

via openrouter

What changed

  • Output limit100K236K2.4×
  • Cache read$0.15/M—
capability
GLM-5.3-Flash

via ofox

What changed

  • open weightsNoYesEnabled
repriced
MiniMax M1

via openrouter

What changed

  • Input price$0.55/M$0.4/M▼ 27.3%
capability
GLM-5.3-Flash

via llmgateway

What changed

  • open weightsNoYesEnabled
repriced
GLM 5.2 TEE

via nano-gpt

What changed

  • Output price$4.60/M$4.40/M▼ 4.3%
  • Cache read$0.5/M$0.7/M▲ 40.0%
repriced
GLM-4.7-Flash

via openrouter

What changed

  • Output limit16K118K7.2×
  • Input price$0.06/M$0.06/M▲ 0.8%
  • Cache read$0.01/M—
new model
Claude Fable 5.1

via vivgrid

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via kilo

What changed

  • open weightsNoYesEnabled
repriced
Kimi K2.5

via hyper

What changed

  • Input price$0.514/M$0.558/M▲ 8.6%
  • Output price$2.75/M$2.94/M▲ 6.5%
  • cache write$0.257/M$0.279/M▲ 8.6%
capability
GLM-5.3-Flash

via scnet-token-plan

What changed

  • open weightsNoYesEnabled
removed
Minimax M2.7 (Free)

via vercel

What changed

Removed from the model catalog

capability
GLM 5.3 Flash

via empiriolabs

What changed

  • open weightsNoYesEnabled
new model
Grok 4.6 (US)

via amazon-bedrock

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via vivgrid

What changed

  • open weightsNoYesEnabled
removed
MiniMax M3 (free)

via openrouter

What changed

Removed from the model catalog

capability
Qwen3.8 27B (Consensus Protocol)

via llmgateway-providers

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
new model
Claude Fable 5

via vivgrid

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via togetherai

What changed

  • open weightsNoYesEnabled
new model
Agentic Chat (GPT-6 Astra)

via gitlab

What changed

First observed in the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.14/M$0.132/M▼ 5.7%
  • Output price$0.58/M$0.528/M▼ 9.0%
  • Cache read$0.035/M$0.033/M▼ 5.7%
capability
GLM 5.3 Flash TEE

via nano-gpt

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3-Flash

via zai-coding-plan

What changed

  • open weightsNoYesEnabled
new model
GLM 5.3 Fast

via fireworks-ai

What changed

First observed in the model catalog

capability
GLM-5.3 Flash (Runware)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
removed
DeepSeek V4 Flash 0731 (FlexAI)

via edenai

What changed

Removed from the model catalog

removed
MiniMax: MiniMax M3 (free)

via kilo

What changed

Removed from the model catalog

capability
Qwen3 30B A3B

via openrouter

What changed

  • structured outputNoYesEnabled
capability
GLM-5.3-Flash

via nan

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3 Flash

via neon

What changed

  • open weightsNoYesEnabled

202 events
repriced
GLM-4.6

via kilo

What changed

  • Context203K198K−2%
  • Output limit131K16K−88%
  • Input price$0.5/M$0.43/M▼ 14.0%
  • Output price$2/M$1.75/M▼ 12.5%
  • +1 more changes
capability
Claude Fable Latest (Claude Fable 5.1)

via edenai

What changed

  • nameClaude Fable 5.1Claude Fable Latest (Claude Fable 5.1)
capability
DeepSeek V4 Flash 0731 (FlexAI)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (FlexAI)
capability
Nemotron 3 Ultra 550B A55B (Nebius)

via edenai

What changed

  • nameNemotron 3 Ultra 550B A55BNemotron 3 Ultra 550B A55B (Nebius)
context
Qwen3.5 397B-A17B

via kilo

What changed

  • Output limit236K66K−72%
capability
GPT OSS 120B (Groq)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Groq)
capability
Gemini 3.8 Flash (Vertex AI)

via edenai

What changed

  • nameGemini 3.8 FlashGemini 3.8 Flash (Vertex AI)
capability
Gemini 3.7 Flash (Vertex AI, US)

via edenai

What changed

  • nameGemini 3.7 Flash (US)Gemini 3.7 Flash (Vertex AI, US)
capability
Nano Banana 2 Preview

via fastrouter

What changed

  • nameNano Banana 2Nano Banana 2 Preview
removed
Kimi K2.5

via moonshotai

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731

via deepinfra

What changed

  • Input price$0.08/M$0.06/M▼ 25.0%
  • Cache read$0.016/M$0.015/M▼ 6.3%
repriced
GPT-5.6 Terra

via venice

What changed

  • Input price$3.13/M$2.50/M▼ 20.0%
  • Output price$18.75/M$15/M▼ 20.0%
  • Cache read$0.313/M$0.25/M▼ 20.0%
  • cache write$3.91/M$3.13/M▼ 20.0%
  • +1 more changes
capability
Nemotron 3 Super 120B A12B (Nebius)

via edenai

What changed

  • nameNemotron 3 Super 120B A12BNemotron 3 Super 120B A12B (Nebius)
capability
Kimi K2.5 (Amazon Bedrock)

via edenai

What changed

  • nameKimi K2.5Kimi K2.5 (Amazon Bedrock)
capability
Inkling Small (Deep Infra)

via edenai

What changed

  • nameInkling SmallInkling Small (Deep Infra)
capability
Nano Banana 2 Lite (Vertex AI)

via edenai

What changed

  • nameNano Banana 2 LiteNano Banana 2 Lite (Vertex AI)
capability
GPT-6 Astra (Azure)

via llmgateway-providers

What changed

  • knowledge—2026-04-30
capability
GPT 6 Astra

via nano-gpt

What changed

  • knowledge—2026-04-30
capability
GPT Latest (GPT-6 Astra)

via edenai

What changed

  • nameGPT-6 AstraGPT Latest (GPT-6 Astra)
  • knowledge—2026-04-30
capability
GPT-5.2 Codex (Azure)

via edenai

What changed

  • nameGPT-5.2 CodexGPT-5.2 Codex (Azure)
capability
Nano Banana Pro Preview

via abacus

What changed

  • nameNano Banana ProNano Banana Pro Preview
capability
GPT OSS 120B (OVHcloud)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (OVHcloud)
capability
DeepSeek V4 Pro 0813 (Together AI)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (Together AI)
capability
DeepSeek V4 Pro 0813 (Databricks)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (Databricks)
capability
GPT Mini Latest (GPT-5.4 mini)

via edenai

What changed

  • nameGPT-5.4 miniGPT Mini Latest (GPT-5.4 mini)
capability
Muse Glimmer 30B (Together AI)

via edenai

What changed

  • nameMuse Glimmer 30BMuse Glimmer 30B (Together AI)
capability
DeepSeek-R1 (Deep Infra)

via edenai

What changed

  • nameDeepSeek-R1DeepSeek-R1 (Deep Infra)
capability
DeepSeek V4 Flash 0731 (TensorX)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (TensorX)
new model
DeepSeek V4 Pro

via sensenova

What changed

First observed in the model catalog

capability
Nano Banana (Vertex AI)

via edenai

What changed

  • nameNano BananaNano Banana (Vertex AI)
capability
GLM-4.7-Flash (Amazon Bedrock, US)

via edenai

What changed

  • nameGLM-4.7-Flash (US)GLM-4.7-Flash (Amazon Bedrock, US)
capability
Llama-Guard-3-8B (Cloudflare)

via edenai

What changed

  • nameLlama-Guard-3-8BLlama-Guard-3-8B (Cloudflare)
capability
GPT OSS 120B (Deep Infra)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Deep Infra)
capability
Gemini 3.7 Flash (Vertex AI)

via edenai

What changed

  • nameGemini 3.7 FlashGemini 3.7 Flash (Vertex AI)
capability
Step 3.7 Flash (Deep Infra)

via edenai

What changed

  • nameStep 3.7 FlashStep 3.7 Flash (Deep Infra)
capability
DeepSeek V4 Pro 0813 (Alibaba)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (Alibaba)
capability
Gemini 3.5 Flash (Vertex AI, US)

via edenai

What changed

  • nameGemini 3.5 Flash (US)Gemini 3.5 Flash (Vertex AI, US)
capability
GPT-5.1 Codex mini (Azure)

via edenai

What changed

  • nameGPT-5.1 Codex miniGPT-5.1 Codex mini (Azure)
capability
Llama-3.3-70B-Instruct (Nebius)

via edenai

What changed

  • nameLlama-3.3-70B-InstructLlama-3.3-70B-Instruct (Nebius)
capability
GPT OSS 20B (Databricks, EU)

via edenai

What changed

  • nameGPT OSS 20B (EU)GPT OSS 20B (Databricks, EU)
capability
GPT OSS 20B (Together AI)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (Together AI)
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit131K944K7.2×
  • Input price$0.05/M$0.045/M▼ 10.0%
  • Output price$0.1/M$0.09/M▼ 10.0%
  • Cache read$0.0100/M$0.0090/M▼ 10.0%
removed
Kimi K2 Thinking Turbo

via moonshotai-cn

What changed

Removed from the model catalog

capability
Llama-Guard-3-8B (Deep Infra)

via edenai

What changed

  • nameLlama-Guard-3-8BLlama-Guard-3-8B (Deep Infra)
capability
GPT-6 Astra

via edenai

What changed

  • knowledge—2026-04-30
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.779/M$1.04/M▲ 32.9%
  • Output price$1.56/M$2.07/M▲ 32.9%
  • Cache read$0.065/M$0.086/M▲ 32.9%
new model
DeepSeek V4 Flash Vision Exp

via crof

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
capability
Seed 2.0 Mini (Deep Infra)

via edenai

What changed

  • nameSeed 2.0 MiniSeed 2.0 Mini (Deep Infra)
capability
GLM-4.7-Flash (Amazon Bedrock)

via edenai

What changed

  • nameGLM-4.7-FlashGLM-4.7-Flash (Amazon Bedrock)
context
GPT-5.1 (2025-11-13)

via nano-gpt

What changed

  • Context1M400K−60%
  • Output limit33K128K3.9×
  • Input limit1M400K−60%
new model
GPT-6 Astra

via neon

What changed

First observed in the model catalog

capability
Gemma-SEA-LION-v4-27B-IT (Cloudflare)

via edenai

What changed

  • nameGemma-SEA-LION-v4-27B-ITGemma-SEA-LION-v4-27B-IT (Cloudflare)
capability
DeepSeek V4 Flash 0731 (Alibaba)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Alibaba)
capability
GPT OSS 120B (Nebius)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Nebius)
capability
Gemini 3.6 Flash (Vertex AI, US)

via edenai

What changed

  • nameGemini 3.6 Flash (US)Gemini 3.6 Flash (Vertex AI, US)
capability
GLM-4.7-Flash (Deep Infra)

via edenai

What changed

  • nameGLM-4.7-FlashGLM-4.7-Flash (Deep Infra)
new model
Kimi K3

via sensenova

What changed

First observed in the model catalog

capability
Nano Banana 2 Preview

via edenai

What changed

  • nameNano Banana 2Nano Banana 2 Preview
capability
Hy3 (Deep Infra)

via edenai

What changed

  • nameHy3Hy3 (Deep Infra)
capability
Nano Banana 2 Preview

via abacus

What changed

  • nameNano Banana 2Nano Banana 2 Preview
capability
GPT OSS 20B (Databricks)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (Databricks)
context
Qwen3-Next 80B-A3B (Thinking)

via kilo

What changed

  • Context131K262K2×
  • Output limit33K236K7.2×
capability
GPT OSS 120B (FlexAI)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (FlexAI)
capability
Gemini 3.5 Flash (Vertex AI)

via edenai

What changed

  • nameGemini 3.5 FlashGemini 3.5 Flash (Vertex AI)
capability
Inkling (Fireworks AI)

via edenai

What changed

  • nameInklingInkling (Fireworks AI)
capability
Seed 2.0 Code (Deep Infra)

via edenai

What changed

  • nameSeed 2.0 CodeSeed 2.0 Code (Deep Infra)
capability
DeepSeek V4 Pro 0813 (Cloudflare)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (Cloudflare)
capability
DeepSeek V4 Flash 0731 (Scaleway)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Scaleway)
capability
GPT OSS 20B (OVHcloud)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (OVHcloud)
capability
GPT OSS 120B (Databricks)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Databricks)
capability
GPT OSS 120B (Databricks, EU)

via edenai

What changed

  • nameGPT OSS 120B (EU)GPT OSS 120B (Databricks, EU)
capability
Nano Banana 2 (Vertex AI)

via edenai

What changed

  • nameNano Banana 2Nano Banana 2 (Vertex AI)
capability
Inkling Small (Together AI)

via edenai

What changed

  • nameInkling SmallInkling Small (Together AI)
capability
GPT-5.1 Codex (Azure)

via edenai

What changed

  • nameGPT-5.1 CodexGPT-5.1 Codex (Azure)
removed
Kimi K2 Thinking Turbo

via moonshotai

What changed

Removed from the model catalog

capability
GLM-4.7-Flash (Cloudflare)

via edenai

What changed

  • nameGLM-4.7-FlashGLM-4.7-Flash (Cloudflare)
capability
Llama-3.2-11B-Vision-Instruct (Deep Infra)

via edenai

What changed

  • nameLlama-3.2-11B-Vision-InstructLlama-3.2-11B-Vision-Instruct (Deep Infra)
capability
Inkling (Deep Infra)

via edenai

What changed

  • nameInklingInkling (Deep Infra)
capability
Gemini 3.5 Flash (Vertex AI, EU)

via edenai

What changed

  • nameGemini 3.5 Flash (EU)Gemini 3.5 Flash (Vertex AI, EU)
capability
Kimi K2.5 (Deep Infra)

via edenai

What changed

  • nameKimi K2.5Kimi K2.5 (Deep Infra)
removed
Kimi K2 0711

via moonshotai-cn

What changed

Removed from the model catalog

capability
gpt-oss-safeguard-120b

via tinfoil

What changed

  • reasoningYesNoRemoved
capability
Muse Glimmer 30B (FlexAI)

via edenai

What changed

  • nameMuse Glimmer 30BMuse Glimmer 30B (FlexAI)
capability
GPT-5.6 Luna

via azure-cognitive-services

What changed

  • statusbeta—
capability
Nano Banana Pro Preview

via edenai

What changed

  • nameNano Banana ProNano Banana Pro Preview
capability
DeepSeek V4 Flash 0731 (Databricks)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Databricks)
capability
Grok Latest (Grok 4.6)

via edenai

What changed

  • nameGrok 4.6Grok Latest (Grok 4.6)
capability
Claude Sonnet 5

via azure-cognitive-services

What changed

  • statusbeta—
removed
Kimi K2 0711

via moonshotai

What changed

Removed from the model catalog

removed
Kimi K2.5

via moonshotai-cn

What changed

Removed from the model catalog

capability
Gemini 3.5 Flash Lite (Vertex AI)

via edenai

What changed

  • nameGemini 3.5 Flash LiteGemini 3.5 Flash Lite (Vertex AI)
repriced
GPT-5.6 Sol

via venice

What changed

  • Input price$6.25/M$2.50/M▼ 60.0%
  • Output price$37.50/M$12.50/M▼ 66.7%
  • Cache read$0.625/M$0.25/M▼ 60.0%
  • cache write$7.81/M$3.13/M▼ 60.0%
  • +1 more changes
capability
Grok 4.1 Fast (Non-Reasoning)

via azure

What changed

  • statusbeta—
capability
DeepSeek V4 Pro 0813 (Fireworks AI)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (Fireworks AI)
context
GLM-5.3

via kilo

What changed

  • Output limit262K944K3.6×
repriced
DeepSeek V4 Flash 0731 (Deep Infra)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Deep Infra)
  • Input price$0.08/M$0.06/M▼ 25.0%
  • Cache read$0.016/M$0.015/M▼ 6.3%
capability
Kimi K2.5 (TensorX)

via edenai

What changed

  • nameKimi K2.5Kimi K2.5 (TensorX)
capability
Gemini 3.6 Flash (Vertex AI, EU)

via edenai

What changed

  • nameGemini 3.6 Flash (EU)Gemini 3.6 Flash (Vertex AI, EU)
new model
GPT-6 Astra

via azure

What changed

First observed in the model catalog

capability
GPT-6 Astra

via openrouter

What changed

  • knowledge—2026-04-30
capability
GPT-5.6 Sol

via azure-cognitive-services

What changed

  • statusbeta—
removed
Kimi K2 Thinking

via moonshotai-cn

What changed

Removed from the model catalog

capability
DeepSeek V4 Flash 0731 (Together AI)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Together AI)
capability
Grok 4.1 Fast (Reasoning)

via azure

What changed

  • statusbeta—
capability
GPT-6 Astra

via kilo

What changed

  • knowledge—2026-04-30
capability
DeepSeek V4 Pro 0813 (TensorX)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (TensorX)
context
GPT 5.5

via nano-gpt

What changed

  • Context1M1.05M1.1×
  • Input limit1M1.05M1.1×
capability
Gemini Pro Latest (Gemini 3.1 Pro Preview, Vertex AI)

via edenai

What changed

  • nameGemini 3.1 Pro PreviewGemini Pro Latest (Gemini 3.1 Pro Preview, Vertex AI)
capability
Muse Glimmer 30B (Deep Infra)

via edenai

What changed

  • nameMuse Glimmer 30BMuse Glimmer 30B (Deep Infra)
capability
Qwen2.5-Coder-32B-Instruct (Cloudflare)

via edenai

What changed

  • nameQwen2.5-Coder-32B-InstructQwen2.5-Coder-32B-Instruct (Cloudflare)
capability
GPT OSS 20B (Cloudflare)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (Cloudflare)
context
Qwen3-Next 80B-A3B (Thinking)

via openrouter

What changed

  • Output limit33K236K7.2×
capability
Nemotron 3 Ultra 550B A55B (Deep Infra)

via edenai

What changed

  • nameNemotron 3 Ultra 550B A55BNemotron 3 Ultra 550B A55B (Deep Infra)
capability
DeepSeek-V3 (Deep Infra)

via edenai

What changed

  • nameDeepSeek-V3DeepSeek-V3 (Deep Infra)
capability
Claude Sonnet 5

via azure

What changed

  • statusbeta—
capability
Nano Banana Pro Preview

via openrouter

What changed

  • nameNano Banana ProNano Banana Pro Preview
new model
Qwen3.8 27B

via cerebras

What changed

First observed in the model catalog

capability
GPT OSS 120B (IONOS)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (IONOS)
capability
Gemini 3.6 Flash (Vertex AI)

via edenai

What changed

  • nameGemini 3.6 FlashGemini 3.6 Flash (Vertex AI)
capability
Gemini 3.1 Pro Preview (Vertex AI)

via edenai

What changed

  • nameGemini 3.1 Pro PreviewGemini 3.1 Pro Preview (Vertex AI)
capability
GPT OSS 120B (Fireworks AI)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Fireworks AI)
capability
Gemini 3.8 Flash (Vertex AI, EU)

via edenai

What changed

  • nameGemini 3.8 Flash (EU)Gemini 3.8 Flash (Vertex AI, EU)
context
GPT 5.4

via nano-gpt

What changed

  • Context922K1.05M1.1×
  • Input limit922K1.05M1.1×
new model
GLM-5.3 Flash

via neon

What changed

First observed in the model catalog

capability
GPT-6 Astra

via merge-gateway

What changed

  • knowledge—2026-04-30
capability
Llama-3.3-70B-Instruct (Deep Infra)

via edenai

What changed

  • nameLlama-3.3-70B-InstructLlama-3.3-70B-Instruct (Deep Infra)
capability
Gemini 3.5 Flash Lite (Vertex AI, US)

via edenai

What changed

  • nameGemini 3.5 Flash Lite (US)Gemini 3.5 Flash Lite (Vertex AI, US)
capability
GPT-6 Astra

via github-copilot

What changed

  • knowledge—2026-04-30
capability
Claude Sonnet Latest (Claude Sonnet 5)

via edenai

What changed

  • nameClaude Sonnet 5Claude Sonnet Latest (Claude Sonnet 5)
capability
GPT-6 Astra

via crossmodel

What changed

  • knowledge—2026-04-30
new model
GPT-6 Astra

via ofox

What changed

First observed in the model catalog

capability
GPT-6 Astra

via venice

What changed

  • knowledge—2026-04-30
capability
GPT-5.6 Terra

via azure

What changed

  • statusbeta—
new model
Grok 4.6

via neon

What changed

First observed in the model catalog

capability
Muse Glimmer 30B (Fireworks AI)

via edenai

What changed

  • nameMuse Glimmer 30BMuse Glimmer 30B (Fireworks AI)
capability
Llama-3.3-70B-Instruct (IONOS)

via edenai

What changed

  • nameLlama-3.3-70B-InstructLlama-3.3-70B-Instruct (IONOS)
capability
Llama-3.3-70B-Instruct (Scaleway)

via edenai

What changed

  • nameLlama-3.3-70B-InstructLlama-3.3-70B-Instruct (Scaleway)
capability
Gemini 3.1 Flash Lite (Vertex AI)

via edenai

What changed

  • nameGemini 3.1 Flash LiteGemini 3.1 Flash Lite (Vertex AI)
capability
GPT OSS 120B (Scaleway)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Scaleway)
capability
Step 3.5 Flash (Deep Infra)

via edenai

What changed

  • nameStep 3.5 FlashStep 3.5 Flash (Deep Infra)
capability
Nemotron 3 Super 120B A12B (FlexAI)

via edenai

What changed

  • nameNemotron 3 Super 120B A12BNemotron 3 Super 120B A12B (FlexAI)
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.082/M$0.09/M▲ 9.0%
  • Output price$0.165/M$0.18/M▲ 9.0%
  • Cache read$0.016/M$0.018/M▲ 9.0%
new model
Inkling (Databricks)

via edenai

What changed

First observed in the model catalog

capability
Gemini Flash Latest (Gemini 3.8 Flash, Vertex AI)

via edenai

What changed

  • nameGemini 3.8 FlashGemini Flash Latest (Gemini 3.8 Flash, Vertex AI)
capability
Inkling (Together AI)

via edenai

What changed

  • nameInklingInkling (Together AI)
new model
TheDrummer/Artemis v1.1

via nano-gpt

What changed

First observed in the model catalog

repriced
GPT-5.6 Sol Pro

via venice

What changed

  • Input price$6.25/M$2.50/M▼ 60.0%
  • Output price$37.50/M$12.50/M▼ 66.7%
  • Cache read$0.625/M$0.25/M▼ 60.0%
  • cache write$7.81/M$3.13/M▼ 60.0%
  • +1 more changes
capability
Nano Banana Pro Preview

via kilo

What changed

  • nameNano Banana ProNano Banana Pro Preview
repriced
GPT-5.6 Terra Pro

via venice

What changed

  • Input price$3.13/M$2.50/M▼ 20.0%
  • Output price$18.75/M$15/M▼ 20.0%
  • Cache read$0.313/M$0.25/M▼ 20.0%
  • cache write$3.91/M$3.13/M▼ 20.0%
  • +1 more changes
capability
Claude Opus Latest (Claude Opus 5)

via edenai

What changed

  • nameClaude Opus 5Claude Opus Latest (Claude Opus 5)
repriced
GLM-5.3

via openrouter

What changed

  • Output limit262K944K3.6×
  • Cache read$0.14/M$0.26/M▲ 85.7%
removed
Gemma 4 31B IT

via cerebras

What changed

Removed from the model catalog

capability
GPT OSS 20B (Groq)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (Groq)
capability
DeepSeek V4 Pro 0813 (Deep Infra)

via edenai

What changed

  • nameDeepSeek V4 Pro 0813DeepSeek V4 Pro 0813 (Deep Infra)
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Output limit944K131K−86%
  • Input price$0.065/M$0.14/M▲ 115.4%
  • Output price$0.18/M$0.28/M▲ 55.6%
  • Cache read$0.016/M$0.028/M▲ 75.0%
capability
GPT-6 Astra

via opencode

What changed

  • knowledge—2026-04-30
capability
Nano Banana Pro (Vertex AI)

via edenai

What changed

  • nameNano Banana ProNano Banana Pro (Vertex AI)
capability
Gemini Pro Latest (Gemini 3.1 Pro Preview)

via edenai

What changed

  • nameGemini 3.1 Pro PreviewGemini Pro Latest (Gemini 3.1 Pro Preview)
capability
Kimi K2 Thinking (Amazon Bedrock)

via edenai

What changed

  • nameKimi K2 ThinkingKimi K2 Thinking (Amazon Bedrock)
capability
GPT OSS 120B (Together AI)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Together AI)
capability
Gemini 3.5 Flash Lite (Vertex AI, EU)

via edenai

What changed

  • nameGemini 3.5 Flash Lite (EU)Gemini 3.5 Flash Lite (Vertex AI, EU)
capability
Nano Banana Pro Preview

via fastrouter

What changed

  • nameNano Banana ProNano Banana Pro Preview
removed
Kimi K2 0905

via moonshotai-cn

What changed

Removed from the model catalog

capability
DeepSeek V3 0324 (Deep Infra)

via edenai

What changed

  • nameDeepSeek V3 0324DeepSeek V3 0324 (Deep Infra)
capability
GPT-5.6 Terra

via azure-cognitive-services

What changed

  • statusbeta—
capability
Gemini 3 Flash Preview (Vertex AI)

via edenai

What changed

  • nameGemini 3 Flash PreviewGemini 3 Flash Preview (Vertex AI)
capability
GPT OSS 20B (Deep Infra)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (Deep Infra)
removed
Kimi K2.5 (DeepInfra)

via llmgateway-providers

What changed

Removed from the model catalog

capability
GPT-6 Astra

via openai

What changed

  • knowledge—2026-04-30
capability
GPT-5.6 Sol

via azure

What changed

  • statusbeta—
capability
DeepSeek V4 Flash 0731 (Cloudflare)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Cloudflare)
capability
Step 3.7 Flash (FlexAI)

via edenai

What changed

  • nameStep 3.7 FlashStep 3.7 Flash (FlexAI)
removed
GLM 5.2 (free)

via openrouter

What changed

Removed from the model catalog

removed
Kimi K2 Turbo

via moonshotai

What changed

Removed from the model catalog

capability
GPT OSS 20B (FlexAI)

via edenai

What changed

  • nameGPT OSS 20BGPT OSS 20B (FlexAI)
capability
Gemini 3.7 Flash (Vertex AI, EU)

via edenai

What changed

  • nameGemini 3.7 Flash (EU)Gemini 3.7 Flash (Vertex AI, EU)
capability
GPT OSS 120B (Cloudflare)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Cloudflare)
capability
GPT-5.6 Luna

via azure

What changed

  • statusbeta—
capability
Gemini 3.1 Flash Lite (Vertex AI, US)

via edenai

What changed

  • nameGemini 3.1 Flash Lite (US)Gemini 3.1 Flash Lite (Vertex AI, US)
capability
GPT Pro Latest (GPT-5.5 Pro)

via edenai

What changed

  • nameGPT-5.5 ProGPT Pro Latest (GPT-5.5 Pro)
capability
Nano Banana 2 Preview

via openrouter

What changed

  • nameNano Banana 2Nano Banana 2 Preview
capability
Nemotron 3 Nano 30B A3B (Deep Infra)

via edenai

What changed

  • nameNemotron 3 Nano 30B A3BNemotron 3 Nano 30B A3B (Deep Infra)
capability
GPT-6 Astra

via vercel

What changed

  • knowledge—2026-04-30
repriced
GLM-4.6

via openrouter

What changed

  • Output limit131K16K−88%
  • Input price$0.5/M$0.43/M▼ 14.0%
  • Output price$2/M$1.75/M▼ 12.5%
  • Cache read$0.1/M$0.08/M▼ 20.0%
capability
Gemini Flash Latest (Gemini 3.8 Flash)

via edenai

What changed

  • nameGemini 3.8 FlashGemini Flash Latest (Gemini 3.8 Flash)
capability
GPT OSS 120B (Cerebras)

via edenai

What changed

  • nameGPT OSS 120BGPT OSS 120B (Cerebras)
repriced
Qwen3.5 397B-A17B

via openrouter

What changed

  • Output limit236K66K−72%
  • Input price$0.55/M$0.39/M▼ 29.1%
  • Output price$3.50/M$2.34/M▼ 33.1%
  • Cache read$0.225/M—
capability
Nano Banana 2 Preview

via kilo

What changed

  • nameNano Banana 2Nano Banana 2 Preview
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit131K944K7.2×
  • Input price$0.05/M$0.045/M▼ 10.0%
  • Output price$0.1/M$0.09/M▼ 10.0%
  • Cache read$0.0100/M$0.0090/M▼ 10.0%
capability
GPT-5.1 Codex Max (Azure)

via edenai

What changed

  • nameGPT-5.1 Codex MaxGPT-5.1 Codex Max (Azure)
capability
DeepSeek V4 Flash 0731 (Nebius)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Nebius)
capability
DeepSeek V4 Flash 0731 (Fireworks AI)

via edenai

What changed

  • nameDeepSeek V4 Flash 0731DeepSeek V4 Flash 0731 (Fireworks AI)
capability
Llama 3.1 Nemotron 70B Instruct (Deep Infra)

via edenai

What changed

  • nameLlama 3.1 Nemotron 70B InstructLlama 3.1 Nemotron 70B Instruct (Deep Infra)
capability
Gemini 3.8 Flash (Vertex AI, US)

via edenai

What changed

  • nameGemini 3.8 Flash (US)Gemini 3.8 Flash (Vertex AI, US)
new model
Claude Fable 5.1

via neon

What changed

First observed in the model catalog

context
DeepSeek V4 Flash 0731

via kilo

What changed

  • Output limit944K131K−86%
removed
Kimi K2 Thinking

via moonshotai

What changed

Removed from the model catalog

removed
Kimi K2 0905

via moonshotai

What changed

Removed from the model catalog

removed
Kimi K2 Turbo

via moonshotai-cn

What changed

Removed from the model catalog

capability
Gemini 3.1 Flash Lite (Vertex AI, EU)

via edenai

What changed

  • nameGemini 3.1 Flash Lite (EU)Gemini 3.1 Flash Lite (Vertex AI, EU)

16 events
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Input price$1.15/M$1.17/M▲ 1.7%
  • Output price$3.50/M$3.96/M▲ 13.1%
  • Cache read$0.1/M$0.234/M▲ 134.0%
context
TheDrummer: UnslopNemo 12B

via kilo

What changed

  • Context33K1.02M31.3×
  • Output limit26K819K31.3×
repriced
GLM-4.6

via openrouter

What changed

  • Input price$0.55/M$0.5/M▼ 9.1%
  • Output price$2.20/M$2/M▼ 9.1%
  • Cache read$0.11/M$0.1/M▼ 9.1%
repriced
GLM Latest

via openrouter

What changed

  • Input price$1.15/M$1.17/M▲ 1.7%
  • Output price$3.50/M$3.96/M▲ 13.1%
  • Cache read$0.1/M$0.234/M▲ 134.0%
repriced
Qwen3.5 35B-A3B

via openrouter

What changed

  • Input price$0.08/M$0.313/M▲ 290.6%
  • Output price$0.75/M$1.25/M▲ 66.7%
  • Cache read—$0.156/M
repriced
GPT-5.6 Luna Pro

via venice

What changed

  • Input price$1.25/M$0.25/M▼ 80.0%
  • Output price$7.50/M$1.50/M▼ 80.0%
  • Cache read$0.125/M$0.025/M▼ 80.0%
  • cache write$1.56/M$0.313/M▼ 80.0%
  • +1 more changes
repriced
GPT-5.6 Luna

via venice

What changed

  • Input price$0.267/M$0.25/M▼ 6.3%
  • Output price$1.60/M$1.50/M▼ 6.3%
  • Cache read$0.027/M$0.025/M▼ 6.3%
  • cache write$0.333/M$0.313/M▼ 6.3%
  • +1 more changes
repriced
GLM-4.6

via kilo

What changed

  • Context205K203K−1%
  • Input price$0.55/M$0.5/M▼ 9.1%
  • Output price$2.20/M$2/M▼ 9.1%
  • Cache read$0.11/M$0.1/M▼ 9.1%
repriced
GPT Latest

via nano-gpt

What changed

  • Input price$2/M$10/M▲ 400.0%
  • Output price$10/M$50/M▲ 400.0%
  • Cache read$0.2/M$1/M▲ 400.0%
  • cache write$2.50/M$12.50/M▲ 400.0%
new model
Inkling

via hyper

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$0.579/M$1.12/M▲ 93.4%
  • Output price$1.74/M$3.36/M▲ 93.4%
  • Cache read$0.019/M$0.037/M▲ 93.4%
context
Qwen3.5 35B-A3B

via kilo

What changed

  • Context262K256K−2%
context
UnslopNemo 12B

via openrouter

What changed

  • Output limit26K819K31.3×
repriced
GLM-5.3 Flash

via merge-gateway

What changed

  • Cache read$0.03/M$0.0030/M▼ 90.0%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.084/M$0.082/M▼ 2.3%
  • Output price$0.169/M$0.165/M▼ 2.3%
  • Cache read$0.017/M$0.016/M▼ 2.3%
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.85/M$0.779/M▼ 8.4%
  • Output price$1.70/M$1.56/M▼ 8.4%
  • Cache read$0.071/M$0.065/M▼ 8.4%

196 events
new model
Claude Fable 5.1

via cloudflare-ai-gateway

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via opencode

What changed

First observed in the model catalog

capability
Qwen Plus

via merge-gateway

What changed

  • structured outputNoYesEnabled
context
Gemini 2.5 Pro

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • structured outputNoYesEnabled
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • +1 more changes
context
Gemini Pro Latest

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
GPT 6 Astra

via nano-gpt

What changed

First observed in the model catalog

new model
GLM 5.3 Fast

via baseten

What changed

First observed in the model catalog

context
Qwen3.5 35B-A3B

via kilo

What changed

  • Output limit236K16K−93%
context
Anubis 70B v1.1

via nano-gpt

What changed

  • Context131K32K−76%
  • Input limit131K32K−76%
capability
MiniMax M3

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
cline-pass/glm-5.3-flash

via cline-pass

What changed

First observed in the model catalog

new model
GLM-5.3 Flash (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

capability
GLM 5.3

via fireworks-ai

What changed

  • last updated2026-08-282026-09-04
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▲ 0.1%
  • Output price$1.05/M$1.05/M▲ 0.1%
context
Grok Imagine Image Quality

via xai

What changed

  • Context8K16K2×
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.089/M$0.084/M▼ 4.7%
  • Output price$0.177/M$0.169/M▼ 4.7%
  • Cache read$0.018/M$0.017/M▼ 4.7%
new model
Nemotron 3 Super 120B A12B

via crusoe

What changed

First observed in the model catalog

capability
Qwen3-VL 235B A22B Instruct

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
Gemini 3.8 Flash

via github-copilot

What changed

First observed in the model catalog

removed
GPT-4.1

via github-copilot

What changed

Removed from the model catalog

new model
Qwen3.8 Max 0902

via openrouter

What changed

First observed in the model catalog

context
Gemini 2.5 Flash Lite Preview

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
inclusionAI: Ling 3.0 Flash Sante (free)

via kilo

What changed

First observed in the model catalog

context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • structured outputNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • +1 more changes
capability
Qwen3.5 Plus

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
Qwen3.6 Plus

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
Qwen3.5 35B-A3B

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.25/M$0.08/M▼ 68.0%
  • Output price$1.25/M$0.75/M▼ 40.0%
  • Cache read$0.25/M—
new model
GPT 6 Astra Pro

via nano-gpt

What changed

First observed in the model catalog

context
Grok Imagine Image

via xai

What changed

  • Context8K16K2×
context
Gemini 2.5 Flash Preview (09/2025) – Thinking

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
context
Grok Imagine Image 2.0

via xai

What changed

  • Context8K64K8×
capability
Qwen3.5 35B A3B

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
MiniMax M2.5

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
GPT-6 Astra

via openai

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via scnet-token-plan

What changed

First observed in the model catalog

capability
Qwen3-Next 80B-A3B Instruct

via merge-gateway

What changed

  • structured outputNoYesEnabled
removed
Claude Sonnet 4.5 (latest)

via github-copilot

What changed

Removed from the model catalog

context
Gemini 2.5 Flash Lite

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
context
Qwen25 VL 72b

via nano-gpt

What changed

  • Context128K32K−75%
  • Input limit128K32K−75%
new model
GPT-6 Astra (Fast)

via vercel

What changed

First observed in the model catalog

new model
Muse Spark 1.2 Contributor (Meta Contributor)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Ling 3.0 Flash Sante (Free)

via vercel

What changed

First observed in the model catalog

capability
Qwen Flash

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
Qwen3.8 Max 0902

via kilo

What changed

First observed in the model catalog

capability
Gemini 2.5 Computer Use Preview (10-2025)

via merge-gateway

What changed

  • structured outputNoYesEnabled
context
Qwen: Qwen3 235B A22B Instruct 2507

via kilo

What changed

  • Output limit236K16K−93%
new model
Ling 3.0 Flash Sante (free)

via openrouter

What changed

First observed in the model catalog

context
Gemini 2.5 Flash Preview (09/2025)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
Claude Fable 5.1

via github-copilot

What changed

First observed in the model catalog

new model
GPT-6 Astra

via github-copilot

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$1.04/M$0.85/M▼ 18.5%
  • Output price$2.08/M$1.70/M▼ 18.5%
  • Cache read$0.087/M$0.071/M▼ 18.5%
capability
MiniMax M2.7

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.418/M$0.484/M▲ 15.8%
  • Output price$1.59/M$1.85/M▲ 16.6%
  • cache write$0.209/M$0.242/M▲ 15.8%
capability
Qwen3.7 Max

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
Qwen3 30B A3B

via merge-gateway

What changed

  • structured outputNoYesEnabled
removed
Granite 4.1 8B

via openrouter

What changed

Removed from the model catalog

context
Qwen3.6 27B

via kilo

What changed

  • Output limit236K66K−72%
removed
GPT-5.2 Codex

via github-copilot

What changed

Removed from the model catalog

repriced
Qwen3 32B

via cortecs

What changed

  • Context40K32K−20%
  • Output limit40K32K−20%
  • Input price$0.099/M$0.089/M▼ 10.1%
  • Output price$0.299/M$0.312/M▲ 4.3%
capability
Qwen3.6 Flash

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
MiniMax M2.7 Highspeed

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
inclusionAI: Ling 3.0 Flash Fin

via kilo

What changed

  • nameLing 3.0 Flash FininclusionAI: Ling 3.0 Flash Fin
new model
Qwen3 235B-A22B Instruct 2507

via crusoe

What changed

First observed in the model catalog

context
Gemini 3.1 Pro (Preview Low)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
capability
Qwen3 Coder Plus

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
GPT-6 Astra Pro

via openrouter

What changed

First observed in the model catalog

new model
GPT-6 Astra

via opencode

What changed

First observed in the model catalog

new model
GPT-6 Astra

via venice

What changed

First observed in the model catalog

capability
Qwen3 Coder Flash

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
Kimi K2.5

via hyper

What changed

  • Input price$0.528/M$0.514/M▼ 2.6%
  • Output price$2.79/M$2.75/M▼ 1.1%
  • cache write$0.264/M$0.257/M▼ 2.6%
context
OpenReasoning Nemotron 32B

via nano-gpt

What changed

  • Context131K33K−75%
  • Input limit131K33K−75%
repriced
DeepSeek V4 Flash (Consensus Protocol)

via llmgateway-providers

What changed

  • Input price$0.13/M$0.05/M▼ 61.5%
  • Output price$0.27/M$0.1/M▼ 63.0%
  • Cache read$0.02/M$0.01/M▼ 50.0%
capability
Kimi K3

via kimi-for-coding

What changed

  • attachmentNoYesEnabled
new model
Qwen3.8 Flash

via scnet-token-plan

What changed

First observed in the model catalog

removed
Claude Opus 4.5 (latest)

via github-copilot

What changed

Removed from the model catalog

new model
Qwen3.8 2.4T A95B

via fireworks-ai

What changed

First observed in the model catalog

capability
Kimi K3-256K

via kimi-for-coding

What changed

  • attachmentNoYesEnabled
repriced
MoonshotAI Kimi Latest

via kilo

What changed

  • Input price$2.50/M$2.55/M▲ 2.0%
  • Output price$14/M$12.75/M▼ 8.9%
  • Cache read$0.29/M$0.256/M▼ 11.7%
capability
Glm 4.5V

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
GLM-5.3

via scnet-token-plan

What changed

First observed in the model catalog

context
Gemini 3 Flash (Preview)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
GLM-5.3 Flash (Consensus Protocol)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
GLM-5.3 Flash

via merge-gateway

What changed

  • structured outputNoYesEnabled
  • Cache read$0.0030/M$0.03/M▲ 900.0%
new model
GPT-6 Astra (OpenAI)

via llmgateway-providers

What changed

First observed in the model catalog

context
Gemini 2.5 Flash (No Thinking)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
context
UnslopNemo 12b v4

via nano-gpt

What changed

  • Context33K8K−75%
  • Input limit33K8K−75%
removed
Hermes 4 70B

via cortecs

What changed

Removed from the model catalog

removed
GPT-5.2

via github-copilot

What changed

Removed from the model catalog

new model
GPT-6 Astra

via openrouter

What changed

First observed in the model catalog

capability
Gemini 3.8 Flash

via edenai

What changed

  • nameGemini 3.7 FlashGemini 3.8 Flash
  • descriptionHigh-efficiency Gemini model for agentic workflows, coding, and multimodal reasoningGoogle's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows
  • knowledge2026-03—
  • release date2026-08-132026-09-02
  • +1 more changes
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit393K131K−67%
  • Input price$0.05/M$0.05/M▼ 0.0%
  • Output price$0.16/M$0.1/M▼ 37.5%
  • Cache read$0.013/M$0.0100/M▼ 23.1%
new model
Muse Spark 1.2 Contributor

via llmgateway

What changed

First observed in the model catalog

removed
Claude Opus 4.6

via github-copilot

What changed

Removed from the model catalog

removed
Qwen3.8 Max

via kilo

What changed

Removed from the model catalog

removed
Granite 4.1 8B

via nano-gpt

What changed

Removed from the model catalog

repriced
Gemini 3.8 Flash (US)

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
repriced
GPT 5.6 Luna (Fast)

via vercel

What changed

  • cache write$0.25/M$0.5/M▲ 100.0%
context
Gemini 2.5 Pro Experimental 0325

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
repriced
Qwen3 235B A22B Instruct 2507

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.087/M$0.09/M▲ 2.9%
  • Output price$0.35/M$0.55/M▲ 57.1%
  • Cache read$0.018/M—
repriced
GPT-5.6 Sol

via crossmodel

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
  • +1 more changes
repriced
Gemini 3.6 Flash

via crossmodel

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$1.50/M$0.75/M▼ 50.0%
repriced
GPT-5.6 Sol

via github-copilot

What changed

  • Input price$2/M$4/M▲ 100.0%
  • Output price$10/M$20/M▲ 100.0%
  • Cache read$0.2/M$0.4/M▲ 100.0%
  • cache write$2.50/M$5/M▲ 100.0%
  • +1 more changes
capability
GPT-4 Turbo

via merge-gateway

What changed

  • structured outputNoYesEnabled
removed
Evayale 70b

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 5.3 Flash

via wandb

What changed

First observed in the model catalog

context
Qwen 3 235b A22B

via nano-gpt

What changed

  • Context41K262K6.4×
  • Output limit33K16K−50%
  • Input limit41K262K6.4×
new model
Qwen3.8 Max 0902

via edenai

What changed

First observed in the model catalog

capability
Ling 3.0 Flash

via openrouter

What changed

  • nameLing-3.0-flashLing 3.0 Flash
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$1.12/M$0.579/M▼ 48.0%
  • Output price$3.35/M$1.74/M▼ 48.0%
  • Cache read$0.037/M$0.019/M▼ 48.0%
repriced
GPT 5.6 Terra (Fast)

via vercel

What changed

  • cache write$2.50/M$5/M▲ 100.0%
context
Claude 4.5 Opus

via nano-gpt

What changed

  • Output limit32K64K2×
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.44/M$0.22/M▼ 50.0%
  • Output price$1.32/M$0.66/M▼ 50.0%
  • Cache read$0.014/M$0.0070/M▼ 50.0%
repriced
GPT 5.6 Sol (Fast)

via vercel

What changed

  • cache write$2.50/M$5/M▲ 100.0%
capability
DeepSeek V4 Pro 0423

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
MiniMax M2

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
Inkling

via edenai

What changed

First observed in the model catalog

capability
MiniMax M2.5 Highspeed

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
Nemotron 3 Nano 30B A3B

via crusoe

What changed

First observed in the model catalog

context
Gemini 2.5 Flash Lite Preview (09/2025)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
capability
Qwen3 Max

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
GPT-6 Astra

via edenai

What changed

First observed in the model catalog

capability
inclusionAI: Ling 3.0 Flash Fin (free)

via kilo

What changed

  • nameLing 3.0 Flash Fin (free)inclusionAI: Ling 3.0 Flash Fin (free)
repriced
GLM-5.3-Flash

via llmgateway

What changed

  • Input price$0.13/M$0.1/M▼ 23.1%
  • Output price$0.4/M$0.25/M▼ 37.5%
  • Cache read$0.024/M$0.02/M▼ 16.7%
capability
DeepSeek V4 Pro 0813

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
qwen2.5-vl-72b-instruct

via cortecs

What changed

  • Input price$0.25/M$1.01/M▲ 305.6%
  • Output price$0.747/M$1.01/M▲ 35.7%
capability
Qwen3.6 35B A3B

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
MiniMax M2.1

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.44/M$0.22/M▼ 50.0%
  • Output price$1.32/M$0.66/M▼ 50.0%
  • Cache read$0.014/M$0.0070/M▼ 50.0%
new model
GPT-6 Astra

via vercel

What changed

First observed in the model catalog

removed
Qwen3.8 Max

via openrouter

What changed

Removed from the model catalog

capability
inclusionAI: Ling 3.0 Flash

via kilo

What changed

  • nameLing-3.0-flashinclusionAI: Ling 3.0 Flash
repriced
GPT-6 Astra

via edenai

What changed

  • nameGPT-5.6 SolGPT-6 Astra
  • descriptionFrontier GPT-5.6 model for complex professional work, coding, and agentic workflowsGPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.
  • familygpt-solgpt-astra
  • knowledge2026-02-16—
  • +7 more changes
new model
Kimi K3 (Runware)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Qwen3.6 27B

via openrouter

What changed

  • Output limit236K66K−72%
  • Input price$0.6/M$0.3/M▼ 50.0%
  • Output price$3.60/M$2/M▼ 44.4%
  • Cache read$0.12/M$0.03/M▼ 75.0%
repriced
Gemini 3.8 Flash

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▲ 0.1%
  • Output price$0.755/M$0.755/M▲ 0.1%
context
Gemini 2.5 Pro Preview 0605

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
GLM-5.3

via opencode

What changed

First observed in the model catalog

capability
Qwen3 235B A22B

via merge-gateway

What changed

  • structured outputNoYesEnabled
context
Gemini 2.5 Pro Preview 0506

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
repriced
GLM-5

via hyper

What changed

  • Input price$0.93/M$0.86/M▼ 7.5%
  • Output price$2.88/M$2.78/M▼ 3.3%
  • cache write$0.465/M$0.43/M▼ 7.5%
context
Gemini 2.5 Flash

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
GPT-6 Astra

via crossmodel

What changed

First observed in the model catalog

new model
GPT-6 Astra (Azure)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
MoonshotAI Kimi Latest

via openrouter

What changed

  • Input price$2.50/M$2.55/M▲ 2.0%
  • Output price$14/M$12.75/M▼ 8.9%
  • Cache read$0.29/M$0.256/M▼ 11.7%
capability
GLM 5.3 Flash

via fireworks-ai

What changed

  • last updated2026-08-262026-09-04
removed
IBM: Granite 4.1 8B

via kilo

What changed

Removed from the model catalog

context
Gemini 3.1 Pro (Preview Custom Tools)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
repriced
Gemini 3.8 Flash

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
context
Gemini 3 Flash Thinking

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
Muse Spark 1.3

via opencode

What changed

First observed in the model catalog

context
Claude 4.5 Opus Thinking

via nano-gpt

What changed

  • Output limit32K64K2×
context
Gemini 3.1 Pro (Preview High)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
GPT-6 Astra

via kilo

What changed

First observed in the model catalog

context
Gemini 3.1 Pro (Preview)

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
capability
Qwen3.8 Flash Next

via amd

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
repriced
Gemini 3.8 Flash (EU)

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
capability
Qwen3.8 Max

via merge-gateway

What changed

  • structured outputNoYesEnabled
context
Gemini 2.5 Flash Preview

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.465/M$0.465/M▲ 0.1%
  • Output price$0.929/M$0.93/M▲ 0.1%
new model
OpenAI: GPT-6 Astra Pro ($$$$)

via kilo

What changed

First observed in the model catalog

new model
GPT-6 Astra

via llmgateway

What changed

First observed in the model catalog

context
Gemini 2.5 Flash Preview Thinking

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
GPT-6 Astra

via merge-gateway

What changed

First observed in the model catalog

capability
DeepSeek V4 Flash 0731

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
DeepSeek V4 Flash

via llmgateway

What changed

  • Input price$0.051/M$0.05/M▼ 2.0%
  • Output price$0.104/M$0.1/M▼ 3.8%
  • Cache read$0.0097/M$0.01/M▲ 3.1%
capability
Qwen3-VL 235B A22B Thinking

via merge-gateway

What changed

  • structured outputNoYesEnabled
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.29/M$1.26/M▼ 2.2%
  • Output price$4.22/M$4.13/M▼ 2.1%
  • cache write$0.645/M$0.631/M▼ 2.2%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▲ 0.1%
  • Output price$0.697/M$0.697/M▲ 0.1%
capability
GPT-4o (2024-05-13)

via merge-gateway

What changed

  • structured outputNoYesEnabled
removed
GLM-4.7

via cortecs

What changed

Removed from the model catalog

new model
Ling 3.0 Flash Sante

via vercel

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro

via merge-gateway

What changed

  • structured outputNoYesEnabled
new model
GPT-6 Astra Pro

via venice

What changed

First observed in the model catalog

removed
Claude Sonnet 4 (latest)

via github-copilot

What changed

Removed from the model catalog

new model
DeepSeek V4 Flash Vision Exp

via opencode

What changed

First observed in the model catalog

context
Gemini 2.5 Pro Preview 0325

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
new model
Qwen3 8B TEE

via nano-gpt

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.755/M$0.755/M▲ 0.1%
  • Output price$0.755/M$0.755/M▲ 0.1%
new model
Gemini 3.8 Flash (Google AI Studio)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Qwen3.7 Plus

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
Qwen3-VL Plus

via merge-gateway

What changed

  • structured outputNoYesEnabled
removed
Gemini 3.1 Pro Preview

via github-copilot

What changed

Removed from the model catalog

context
Gemini 3 Pro Image

via nano-gpt

What changed

  • Context1.05M66K−94%
  • Output limit66K33K−50%
  • Input limit1.05M66K−94%
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit393K131K−67%
  • Input price$0.05/M$0.05/M▼ 0.0%
  • Output price$0.16/M$0.1/M▼ 37.5%
  • Cache read$0.013/M$0.0100/M▼ 23.1%
removed
Hermes 4 (Thinking)

via nano-gpt

What changed

Removed from the model catalog

new model
Gemini 3.8 Flash

via crossmodel

What changed

First observed in the model catalog

capability
Qwen3.6 Max Preview

via merge-gateway

What changed

  • structured outputNoYesEnabled
context
Gemini 2.5 Flash Lite Preview (09/2025) – Thinking

via nano-gpt

What changed

  • Context1.05M1.05M−0%
  • Input limit1.05M1.05M−0%
context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • structured outputNoYesEnabled
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • +1 more changes
capability
DeepSeek V4 Flash

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
Qwen3.5 Flash

via merge-gateway

What changed

  • structured outputNoYesEnabled
capability
Gemini 3.8 Flash

via edenai

What changed

  • nameGemini 3.7 FlashGemini 3.8 Flash
  • descriptionHigh-efficiency Gemini model for agentic workflows, coding, and multimodal reasoningGoogle's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows
  • knowledge2026-03—
  • release date2026-08-132026-09-02
  • +1 more changes
new model
Omen Alpha

via opencode-go

What changed

First observed in the model catalog

capability
GLM-5.3

via merge-gateway

What changed

  • structured outputNoYesEnabled

13 events
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$1.02/M$1.04/M▲ 1.9%
  • Output price$2.04/M$2.08/M▲ 1.9%
  • Cache read$0.085/M$0.087/M▲ 1.9%
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.078/M$0.089/M▲ 13.8%
  • Output price$0.156/M$0.177/M▲ 13.8%
  • Cache read$0.016/M$0.018/M▲ 13.8%
repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
repriced
ReMM SLERP 13B

via openrouter

What changed

  • Output limit4K6K1.3×
  • Input price$0.45/M$0.35/M▼ 22.2%
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new model
Nemotron 3.5 Content Safety

via kilo

What changed

First observed in the model catalog

new model
Nemotron 3.5 Content Safety

via openrouter

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$0.66/M$1.12/M▲ 69.0%
  • Output price$1.98/M$3.35/M▲ 69.0%
  • Cache read$0.022/M$0.037/M▲ 69.0%
repriced
Nemotron 3 Ultra 550B A55B

via openrouter

What changed

  • Output limit183K33K−82%
  • Input price$0.6/M$0.625/M▲ 4.2%
  • Output price$2.40/M$3.13/M▲ 30.2%
  • Cache read$0.12/M$0.188/M▲ 56.3%
context
ReMM SLERP 13B

via kilo

What changed

  • Output limit4K6K1.3×
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
context
Nemotron 3 Ultra 550B A55B

via kilo

What changed

  • Context203K256K1.3×
  • Output limit183K33K−82%

172 events
new model
Gemini 3.8 Flash

via edenai

What changed

First observed in the model catalog

context
Qwen25 VL 72b

via nano-gpt

What changed

  • Context32K128K4×
  • Output limit33K115K3.5×
  • Input limit32K128K4×
repriced
Nemotron 3 Nano 30B A3B

via openrouter

What changed

  • Output limit228K236K1×
  • Cache read$0.025/M$0.03/M▲ 20.0%
new model
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

First observed in the model catalog

repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context262K1.05M4×
  • Output limit236K944K4×
  • Input price$1.15/M$1.09/M▼ 5.0%
  • Output price$3.50/M$3.43/M▼ 1.9%
  • +1 more changes
new model
Muse Spark 1.3

via meta

What changed

First observed in the model catalog

repriced
GLM Latest

via openrouter

What changed

  • Output limit236K944K4×
  • Input price$1.15/M$1.09/M▼ 5.0%
  • Output price$3.50/M$3.43/M▼ 1.9%
  • Cache read$0.1/M$0.179/M▲ 79.4%
context
EVA Llama 3.33 70B

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
new model
Gemini 3.8 Flash (EU)

via edenai

What changed

First observed in the model catalog

context
EVA-LLaMA-3.33-70B-v0.1

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
context
GLM 4.6 Turbo (Thinking)

via nano-gpt

What changed

  • Context200K205K1×
  • Output limit205K131K−36%
  • Input limit200K205K1×
repriced
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit131K262K2×
  • Cache read$0.2/M$0.25/M▲ 25.0%
context
GLM 4.5V Thinking

via nano-gpt

What changed

  • Context64K66K1×
  • Output limit96K16K−83%
  • Input limit64K66K1×
context
Kimi Latest

via nano-gpt

What changed

  • Output limit1.05M944K−10%
repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
context
MS3.2 24B Magnum Diamond

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
new model
Qwen3.8 Max 0902

via empiriolabs

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$0.66/M$1.12/M▲ 69.0%
  • Output price$1.98/M$3.35/M▲ 69.0%
  • Cache read$0.022/M$0.037/M▲ 69.0%
context
Grok Latest

via nano-gpt

What changed

  • Output limit500K450K−10%
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
new model
Qwen3.8 27B

via llmgateway

What changed

First observed in the model catalog

repriced
GLM-5

via hyper

What changed

  • Input price$0.85/M$0.93/M▲ 9.4%
  • Output price$2.77/M$2.88/M▲ 3.7%
  • cache write$0.425/M$0.465/M▲ 9.4%
new model
MiniCPM5-1B

via amd

What changed

First observed in the model catalog

context
Qwen3 VL 235B A22B Instruct

via nano-gpt

What changed

  • Context128K131K1×
  • Output limit262K33K−88%
  • Input limit128K131K1×
new model
Claude Fable 5.1

via ofox

What changed

First observed in the model catalog

context
Grok 4.3

via nano-gpt

What changed

  • Output limit1M900K−10%
new model
Muse Spark 1.3

via empiriolabs

What changed

First observed in the model catalog

repriced
Qwen3.6 35B-A3B

via kilo

What changed

  • Output limit236K16K−93%
  • Input price$0.1/M$0.05/M▼ 50.0%
  • Output price$0.9/M$0.7/M▼ 22.2%
  • Cache read$0.05/M—
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▲ 0.3%
  • Output price$0.695/M$0.697/M▲ 0.3%
capability
Muse Spark 1.3

via kilo

What changed

  • nameMeta: Muse Spark 1.3Muse Spark 1.3
repriced
DeepSeek Chat

via kilo

What changed

  • Context128K164K1.3×
  • Output limit16K16K1×
  • Input price$0.257/M$0.32/M▲ 24.3%
  • Output price$1.03/M$0.89/M▼ 13.5%
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
context
DeepSeek R1 0528

via nano-gpt

What changed

  • Context128K164K1.3×
  • Output limit164K33K−80%
  • Input limit128K164K1.3×
context
Solar Pro 3

via nano-gpt

What changed

  • Context128K131K1×
  • Output limit128K118K−8%
  • Input limit128K131K1×
new model
Muse Spark 1.3 Contributor

via llmgateway

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via deepinfra

What changed

First observed in the model catalog

context
Omega Directive 24B Unslop v2.0

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
context
The Drummer Cydonia 24B v4

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
repriced
DeepSeek V3.1

via openrouter

What changed

  • Output limit33K145K4.4×
  • Input price$0.25/M$0.55/M▲ 120.0%
  • Output price$0.95/M$1.65/M▲ 73.7%
  • Cache read$0.13/M$0.55/M▲ 323.1%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▲ 0.3%
  • Output price$0.753/M$0.755/M▲ 0.3%
capability
Qwen3.8 Max 0902

via nano-gpt

What changed

  • descriptionQwen3.8 Max 0902 is Alibaba's September 2 checkpoint of its flagship Qwen3.8 Max model for coding, knowledge work, data analysis, and long-running agent workflows. It supports text, image, video, PDF input, selectable thinking, tool calling, structured output, and a near-million-token context window.2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding
  • familyqwen3.8-maxqwen
repriced
Llama-3.3-70B-Instruct

via openrouter

What changed

  • Output limit115K16K−86%
  • Input price$0.71/M$0.1/M▼ 85.9%
  • Output price$0.71/M$0.32/M▼ 54.9%
  • Cache read$0.71/M—
capability
Abliterated Model Large V2

via nano-gpt

What changed

  • tool callNoYesEnabled
context
Nvidia Nemotron 3 Nano 30B

via nano-gpt

What changed

  • Context256K262K1×
  • Output limit262K236K−10%
  • Input limit256K262K1×
context
Perplexity Academic Researcher

via nano-gpt

What changed

  • Context127K128K1×
  • Output limit128K115K−10%
  • Input limit127K128K1×
new provider
GLM-5.2

via nan

What changed

nan began listing this model

new model
Qwen3.8 27B (Consensus Protocol)

via llmgateway-providers

What changed

First observed in the model catalog

context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
new model
Qwen3.8 Max 0902

via ofox

What changed

First observed in the model catalog

context
Qwen3.5 35B A3B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
new model
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

First observed in the model catalog

context
Llama 3.3 70B Wayfarer

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
new model
Muse Spark 1.3 Contributor (Meta Contributor)

via llmgateway-providers

What changed

First observed in the model catalog

context
Qwen3.8 2.4T A95B

via kilo

What changed

  • Context262K1M3.8×
  • Output limit131K262K2×
context
Nex N2 Mini

via nano-gpt

What changed

  • Output limit262K236K−10%
repriced
Z.ai: GLM Flash Latest

via kilo

What changed

  • Output limit944K131K−86%
  • Input price$0.075/M$0.071/M▼ 5.0%
  • Output price$0.25/M$0.237/M▼ 5.0%
  • Cache read$0.015/M$0.014/M▼ 5.0%
context
Grok 4.5

via nano-gpt

What changed

  • Output limit500K450K−10%
repriced
Qwen3.8 Flash

via edenai

What changed

  • Input price$0.16/M$0.15/M▼ 6.3%
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new model
GLM-5.3-Flash

via tinfoil

What changed

First observed in the model catalog

context
Perplexity Reasoning Pro

via nano-gpt

What changed

  • Context127K128K1×
  • Output limit128K115K−10%
  • Input limit127K128K1×
capability
Step 3.7 Flash

via edenai

What changed

  • structured outputNoYesEnabled
context
Tencent Hy3

via nano-gpt

What changed

  • Output limit262K128K−51%
context
Llama-3.3-70B-Instruct

via kilo

What changed

  • Context128K131K1×
  • Output limit115K16K−86%
new provider
DeepSeek V4 Flash

via nan

What changed

nan began listing this model

context
Kimi K2 0711 Instruct FP4

via nano-gpt

What changed

  • Context128K131K1×
  • Input limit128K131K1×
capability
Qwen3.8 Max 0902

via vercel

What changed

  • familyqwen3.8-maxqwen
  • release date2026-09-012026-09-02
  • last updated2026-09-012026-09-02
context
Mistral Saba

via nano-gpt

What changed

  • Context32K33K1×
  • Output limit33K26K−20%
  • Input limit32K33K1×
context
Grok Build 0.1

via nano-gpt

What changed

  • Output limit256K230K−10%
context
The Drummer Cydonia 24B v2

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
new model
Ling 3.0 Flash Fin

via openrouter

What changed

First observed in the model catalog

context
GLM 4.5V

via nano-gpt

What changed

  • Context64K66K1×
  • Output limit96K16K−83%
  • Input limit64K66K1×
context
Perplexity Deep Research

via nano-gpt

What changed

  • Context60K128K2.1×
  • Output limit128K115K−10%
  • Input limit60K128K2.1×
new provider
Qwen3.6 35B-A3B

via nan

What changed

nan began listing this model

repriced
GLM-5.3

via openrouter

What changed

  • Output limit131K262K2×
  • Cache read$0.26/M$0.14/M▼ 46.2%
repriced
Qwen3.8 27B

via openrouter

What changed

  • Input price$0.425/M$0.42/M▼ 1.2%
  • Output price$2.55/M$3/M▲ 17.6%
  • cache write$0.531/M—
new model
GLM 5.3 Fast

via vercel

What changed

First observed in the model catalog

repriced
Qwen3.5 397B-A17B

via openrouter

What changed

  • Output limit66K236K3.6×
  • Input price$0.39/M$0.55/M▲ 41.0%
  • Output price$2.34/M$3.50/M▲ 49.6%
  • Cache read—$0.225/M
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.463/M$0.465/M▲ 0.3%
  • Output price$0.926/M$0.929/M▲ 0.3%
new model
DeepSeek V4 Flash Vision Exp

via huggingface

What changed

First observed in the model catalog

context
TheDrummer Skyfall 36B V2

via nano-gpt

What changed

  • Context32K33K1×
  • Output limit33K29K−10%
  • Input limit32K33K1×
context
Muse Spark 1.1

via nano-gpt

What changed

  • Context1M1.05M1×
new model
Muse Spark 1.3 (Meta)

via llmgateway-providers

What changed

First observed in the model catalog

removed
GPT OSS 120B

via edenai

What changed

Removed from the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.04/M$1.05/M▲ 0.3%
  • Output price$1.04/M$1.05/M▲ 0.3%
removed
GLM Z1 Air

via nano-gpt

What changed

Removed from the model catalog

context
GLM-5.3

via kilo

What changed

  • Output limit131K262K2×
context
Step 3.5 Flash

via nano-gpt

What changed

  • Context256K262K1×
  • Output limit256K66K−74%
  • Input limit256K262K1×
repriced
Nemotron 3 Ultra 550B A55B

via openrouter

What changed

  • Output limit183K33K−82%
  • Input price$0.6/M$0.625/M▲ 4.2%
  • Output price$2.40/M$3.13/M▲ 30.2%
  • Cache read$0.12/M$0.188/M▲ 56.3%
context
Muse Spark 1.1

via orcarouter

What changed

  • Context1M1.05M1×
  • Output limit32K131K4.1×
new provider
MiMo-V2.5

via nan

What changed

nan began listing this model

repriced
Qwen2.5 VL 72B Instruct

via openrouter

What changed

  • Output limit29K115K4×
  • Input price$0.25/M$0.8/M▲ 220.0%
  • Output price$0.75/M$1/M▲ 33.3%
  • Cache read—$0.4/M
context
Kimi K3 TEE

via nano-gpt

What changed

  • Output limit1.05M66K−94%
new model
Gemini 3.8 Flash

via edenai

What changed

First observed in the model catalog

context
GLM 4.6 Turbo

via nano-gpt

What changed

  • Context200K205K1×
  • Output limit205K131K−36%
  • Input limit200K205K1×
context
Step 3.5 Flash 2603

via nano-gpt

What changed

  • Context256K262K1×
  • Output limit256K66K−74%
  • Input limit256K262K1×
context
Grayline Qwen3 8B

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
context
UnslopNemo 12b v4

via nano-gpt

What changed

  • Context8K33K4×
  • Output limit8K26K3.2×
  • Input limit8K33K4×
new model
Gemini 3.8 Flash (US)

via edenai

What changed

First observed in the model catalog

context
Nex N2 Pro

via nano-gpt

What changed

  • Output limit262K236K−10%
context
Muse Spark 1.1

via llmgateway

What changed

  • Output limit32K131K4.1×
new provider
Gemma 4 26B A4B IT

via nan

What changed

nan began listing this model

repriced
GLM 5.3

via vercel

What changed

  • Input price$1.40/M$0.7/M▼ 50.0%
  • Output price$4.40/M$2.20/M▼ 50.0%
  • Cache read$0.14/M$0.13/M▼ 7.1%
new model
Muse Spark 1.3

via llmgateway

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via amd

What changed

First observed in the model catalog

context
Kimi K3

via nano-gpt

What changed

  • Output limit1.05M944K−10%
context
The Drummer Cydonia 24B v4.1

via nano-gpt

What changed

  • Output limit131K118K−10%
context
Muse Glimmer 30B

via nano-gpt

What changed

  • Output limit131K118K−10%
context
Steelskull Electra R1 70b

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
context
OpenReasoning Nemotron 32B

via nano-gpt

What changed

  • Context33K131K4×
  • Input limit33K131K4×
context
Llama 3.3 70B Cu Mai

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
repriced
IBM: Granite 4.2 8B

via kilo

What changed

  • Input price$0.1/M$0.06/M▼ 40.0%
  • Output price$0.15/M$0.25/M▲ 66.7%
  • Cache read$0.05/M$0.015/M▼ 70.0%
context
Muse Spark 1.1

via meta

What changed

  • Output limit32K131K4.1×
repriced
Nemotron 3 Nano 30B A3B

via kilo

What changed

  • Output limit228K236K1×
  • Cache read$0.025/M$0.03/M▲ 20.0%
repriced
Qwen3.6 35B-A3B

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.1/M$0.05/M▼ 50.0%
  • Output price$0.9/M$0.7/M▼ 22.2%
  • Cache read$0.05/M—
context
Kimi K2 0905

via nano-gpt

What changed

  • Context256K262K1×
  • Output limit262K100K−62%
  • Input limit256K262K1×
removed
Qwen3.8 27B

via llmgateway

What changed

Removed from the model catalog

repriced
GLM-4.6

via openrouter

What changed

  • Output limit16K131K8×
  • Input price$0.43/M$0.55/M▲ 27.9%
  • Output price$1.75/M$2.20/M▲ 25.7%
  • Cache read$0.08/M$0.11/M▲ 37.5%
repriced
MoonshotAI Kimi Latest

via kilo

What changed

  • Input price$2.55/M$2.50/M▼ 2.0%
  • Output price$12.75/M$14/M▲ 9.8%
  • Cache read$0.256/M$0.29/M▲ 13.3%
context
Mistral Small 3.2 24b Instruct

via nano-gpt

What changed

  • Output limit131K16K−88%
new model
Muse Spark 1.3 Contributor

via meta

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.079/M$0.078/M▼ 1.9%
  • Output price$0.159/M$0.156/M▼ 1.9%
  • Cache read$0.016/M$0.016/M▼ 1.9%
new provider
Qwen3.8 Flash

via nan

What changed

nan began listing this model

context
Llama 3.3 70B Instruct abliterated

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
capability
Muse Spark 1.3

via vercel

What changed

  • structured output—YesEnabled
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.753/M$0.755/M▲ 0.3%
  • Output price$0.753/M$0.755/M▲ 0.3%
context
Qwen: Qwen3 14B

via kilo

What changed

  • Context41K131K3.2×
  • Output limit16K8K−50%
context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
context
MythoMax 13B

via nano-gpt

What changed

  • Context4K4K1×
  • Output limit4K4K−10%
  • Input limit4K4K1×
context
ByteDance Seed 2.1 Turbo

via nano-gpt

What changed

  • Output limit262K236K−10%
context
Muse Glimmer 30B

via kilo

What changed

  • Output limit16K118K7.2×
context
Holo3-35B-A3B Thinking

via nano-gpt

What changed

  • Output limit66K8K−88%
new provider
GLM-5.3-Flash

via nan

What changed

nan began listing this model

context
DeepSeek: DeepSeek V3.1

via kilo

What changed

  • Context164K161K−2%
  • Output limit33K145K4.4×
removed
Dots3-Note Preview

via nano-gpt

What changed

Removed from the model catalog

repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.426/M$0.418/M▼ 1.9%
  • Output price$1.62/M$1.59/M▼ 2.0%
  • cache write$0.213/M$0.209/M▼ 1.9%
removed
Hermes 4 Medium

via nano-gpt

What changed

Removed from the model catalog

context
Manta Pro 1.0

via nano-gpt

What changed

  • Context33K66K2×
  • Input limit33K66K2×
repriced
MoonshotAI Kimi Latest

via openrouter

What changed

  • Input price$2.55/M$2.50/M▼ 2.0%
  • Output price$12.75/M$14/M▲ 9.8%
  • Cache read$0.256/M$0.29/M▲ 13.3%
context
Mixtral 8x22B

via nano-gpt

What changed

  • Output limit66K52K−20%
context
Holo3-35B-A3B

via nano-gpt

What changed

  • Output limit66K8K−88%
repriced
Qwen3 14B

via openrouter

What changed

  • Output limit16K8K−50%
  • Input price$0.12/M$0.228/M▲ 89.6%
  • Output price$0.24/M$0.91/M▲ 279.2%
context
Mistral Large 2411

via nano-gpt

What changed

  • Output limit256K102K−60%
repriced
Muse Glimmer 30B

via openrouter

What changed

  • Output limit16K118K7.2×
  • Output price$1.20/M$1.10/M▼ 8.3%
removed
GLM-5.2

via tinfoil

What changed

Removed from the model catalog

repriced
GLM-4.6

via kilo

What changed

  • Context198K205K1×
  • Output limit16K131K8×
  • Input price$0.43/M$0.55/M▲ 27.9%
  • Output price$1.75/M$2.20/M▲ 25.7%
  • +1 more changes
new model
Ornith 1.5 9B

via nano-gpt

What changed

First observed in the model catalog

repriced
Qwen: Qwen2.5 VL 72B Instruct

via kilo

What changed

  • Context32K128K4×
  • Output limit29K115K4×
  • Input price$0.25/M$0.8/M▲ 220.0%
  • Output price$0.75/M$1/M▲ 33.3%
  • +1 more changes
context
Qwen3.5 397B-A17B

via kilo

What changed

  • Output limit66K236K3.6×
context
Z.ai: GLM Flash Latest

via kilo

What changed

  • Output limit131K944K7.2×
context
DeepSeek V4 Flash TEE

via nano-gpt

What changed

  • Output limit1.05M393K−63%
capability
Muse Spark 1.3

via nano-gpt

What changed

  • descriptionMeta's Muse Spark 1.3 is a frontier multimodal reasoning model for long-horizon coding and agentic workflows, with strong gains in computer use, browsing, professional tool use, codebase understanding, instruction following, and million-token retrieval. It accepts text, images, audio, video, and files, supports tool calling and structured output, and always reasons before answering.Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.
  • modalities.inputtext,image,video,audio,pdftext,image,video,pdf,audio
context
Gemma 3 12B IT

via nano-gpt

What changed

  • Context128K131K1×
  • Output limit131K16K−88%
  • Input limit128K131K1×
new model
Ling 3.0 Flash Fin

via kilo

What changed

First observed in the model catalog

context
Qwen3 Next 80B A3B (Instruct)

via nano-gpt

What changed

  • Context256K262K1×
  • Output limit262K236K−10%
  • Input limit256K262K1×
context
Qwen 3 235b A22B 2507 Thinking

via nano-gpt

What changed

  • Context256K131K−49%
  • Output limit262K118K−55%
  • Input limit256K131K−49%
context
GLM Flash Latest

via openrouter

What changed

  • Output limit131K944K7.2×
repriced
GLM Flash Latest

via openrouter

What changed

  • Output limit944K131K−86%
  • Input price$0.075/M$0.071/M▼ 5.0%
  • Output price$0.25/M$0.237/M▼ 5.0%
  • Cache read$0.015/M$0.014/M▼ 5.0%
repriced
Claude Fable Latest

via nano-gpt

What changed

  • Cache read$1/M$0.25/M▼ 75.0%
context
Granite 4.1 8B

via nano-gpt

What changed

  • Output limit131K118K−10%
context
Steelskull Nevoria 70b

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
context
Mistral Small 31 24b Instruct

via nano-gpt

What changed

  • Output limit131K102K−22%
context
Llama 3.1 70B Celeste v0.1

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
new model
Ornith 1.5 9B Thinking

via nano-gpt

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.116/M$0.106/M▼ 8.6%
  • Output price$0.38/M$0.368/M▼ 3.2%
  • cache write$0.058/M$0.053/M▼ 8.6%
context
Grok 4.6

via nano-gpt

What changed

  • Output limit500K450K−10%
repriced
DeepSeek Chat

via openrouter

What changed

  • Output limit16K16K1×
  • Input price$0.257/M$0.32/M▲ 24.3%
  • Output price$1.03/M$0.89/M▼ 13.5%
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.87/M$1.04/M▲ 19.8%
  • Output price$1.74/M$2.08/M▲ 19.8%
  • Cache read$0.072/M$0.087/M▲ 19.8%
context
Qwen 3 235b A22B 2507

via nano-gpt

What changed

  • Context256K262K1×
  • Output limit262K236K−10%
  • Input limit256K262K1×
context
Perplexity Simple

via nano-gpt

What changed

  • Context127K127K1×
  • Output limit128K114K−11%
  • Input limit127K127K1×
context
Llama 3.1 70B Hanami

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
context
Steelskull Nevoria R1 70b

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×

265 events
context
GLM 5.3 Flash Uncensored

via nano-gpt

What changed

  • Context262K524K2×
  • Input limit262K524K2×
context
DeepSeek V4 Pro

via nebius

What changed

  • Context1M1.05M1×
  • Output limit384K1.05M2.7×
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
context
Hy3

via openrouter

What changed

  • Input limit—192K
removed
Qwen2.5-VL-72B-Instruct

via nebius

What changed

Removed from the model catalog

context
Claude Opus 4.6

via cortecs

What changed

  • Output limit1M128K−87%
context
Hy3

via tencent-tokenhub

What changed

  • Output limit64K128K2×
  • Input limit—192K
new model
Z.ai: GLM Flash Latest

via kilo

What changed

First observed in the model catalog

repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.19/M$0.188/M▼ 1.1%
  • Output price$0.63/M$0.7/M▲ 11.1%
  • cache write$0.095/M$0.094/M▼ 1.1%
removed
Nemotron-3-Nano-Omni

via nebius

What changed

Removed from the model catalog

context
Hy3

via deepinfra

What changed

  • Output limit64K128K2×
  • Input limit—192K
new model
Muse Spark 1.3 Free

via opencode

What changed

First observed in the model catalog

context
Hy3

via llmgateway

What changed

  • Output limit64K128K2×
  • Input limit—192K
context
GPT-4o mini

via cortecs

What changed

  • Output limit128K16K−88%
new model
GLM-5.3-Flash

via nebius

What changed

First observed in the model catalog

repriced
GPT OSS 20B

via llmgateway

What changed

  • Input price$0.05/M$0.04/M▼ 20.0%
  • Output price$0.2/M$0.19/M▼ 5.0%
  • Cache read—$0.01/M
removed
INTELLECT-3

via nebius

What changed

Removed from the model catalog

context
Gemini 2.5 Flash

via cortecs

What changed

  • Output limit1.05M66K−94%
repriced
GPT OSS 120B (EU)

via edenai

What changed

  • Cache read$0.15/M$0.015/M▼ 90.0%
new model
Meta: Muse Spark 1.3 Contributor

via kilo

What changed

First observed in the model catalog

repriced
GLM-5

via hyper

What changed

  • Input price$0.91/M$0.85/M▼ 6.6%
  • Output price$2.93/M$2.77/M▼ 5.5%
  • cache write$0.455/M$0.425/M▼ 6.6%
context
Hy3

via requesty

What changed

  • Input limit—192K
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.464/M$0.463/M▼ 0.1%
  • Output price$0.927/M$0.926/M▼ 0.1%
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.404/M$0.426/M▲ 5.4%
  • Output price$1.50/M$1.62/M▲ 8.3%
  • cache write$0.202/M$0.213/M▲ 5.4%
context
Nemotron-3-Super-120B-A12B

via nebius

What changed

  • Context256K262K1×
  • Input limit256K262K1×
context
mistral-large-2402

via cortecs

What changed

  • Output limit32K8K−74%
context
GLM-4.7

via cortecs

What changed

  • Output limit198K203K1×
removed
Baichuan 4 Turbo

via nano-gpt

What changed

Removed from the model catalog

context
GPT-4.1

via cortecs

What changed

  • Output limit1.05M33K−97%
context
Claude Opus 4.7

via cortecs

What changed

  • Output limit1M128K−87%
context
Hy3

via vercel

What changed

  • Input limit—192K
context
GPT-5.6 Sol

via cortecs

What changed

  • Output limit1.05M128K−88%
repriced
Qwen3Guard-Gen-0.6B

via ovhcloud

What changed

  • Input price—$0/M
  • Output price—$0/M
new model
DeepSeek V4 Pro 0813 (EU)

via requesty

What changed

First observed in the model catalog

context
Mistral Small 4

via cortecs

What changed

  • Output limit262K256K−2%
repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.12/M$0.116/M▼ 3.3%
  • Output price$0.42/M$0.38/M▼ 9.5%
  • cache write$0.06/M$0.058/M▼ 3.3%
removed
Meituan: LongCat 2.0 (free)

via kilo

What changed

Removed from the model catalog

new model
Qwen3.8 Flash Next (EU)

via requesty

What changed

First observed in the model catalog

removed
Cohere North Mini Code 1.0

via nano-gpt

What changed

Removed from the model catalog

context
Qwen3-Coder 30B-A3B Instruct

via cortecs

What changed

  • Output limit262K262K1×
removed
DeepSeek-V3.2

via nebius

What changed

Removed from the model catalog

new model
Gemini 3.8 Flash

via vercel

What changed

First observed in the model catalog

new model
Gemini 3.8 Flash (Google Vertex AI)

via llmgateway-providers

What changed

First observed in the model catalog

context
Qwen3.5 35B A3B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.05M1.02M−2%
new model
Kimi K2 Thinking

via hyper

What changed

First observed in the model catalog

new model
DeepSeek V4 Pro 0813

via edenai

What changed

First observed in the model catalog

context
GPT-5

via cortecs

What changed

  • Output limit400K128K−68%
context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
removed
GLM 4.7 TEE

via nano-gpt

What changed

Removed from the model catalog

context
Claude Opus 5

via cortecs

What changed

  • Output limit1M128K−87%
new model
Qwen3.8 27B

via berget

What changed

First observed in the model catalog

context
Claude Sonnet 4.6

via cortecs

What changed

  • Output limit1M128K−87%
removed
Muse Spark 1.3 Contributor

via opencode-go

What changed

Removed from the model catalog

new model
Muse Spark 1.3 Contributor

via vercel

What changed

First observed in the model catalog

repriced
Inkling

via neon

What changed

  • Input price—$1/M
  • Output price—$4.05/M
  • Cache read—$0.17/M
context
Kimi K2.5

via cortecs

What changed

  • Output limit256K262K1×
context
GLM-5.3-Flash

via requesty

What changed

  • Context131K1.05M8×
repriced
GLM-5.3

via cortecs

What changed

  • Input price$1.75/M$1.40/M▼ 20.0%
  • Output price$4.50/M$4.40/M▼ 2.2%
  • Cache read$0.438/M$0.26/M▼ 40.6%
context
Gemma 4 26B A4B IT

via cortecs

What changed

  • Output limit262K82K−69%
context
DeepSeek V4 Pro

via requesty

What changed

  • Output limit384K131K−66%
context
Muse Spark 1.1

via meta

What changed

  • Context1M1.05M1×
repriced
GLM 5.3 Flash

via fireworks-ai

What changed

  • Cache read$0.029/M$0.03/M▲ 3.4%
new model
Gemini 3.8 Flash (EU)

via requesty

What changed

First observed in the model catalog

removed
Mistral Medium 3.5 128B

via berget

What changed

Removed from the model catalog

context
Qwen3.5 397B-A17B

via cortecs

What changed

  • Output limit250K262K1×
new model
Nemotron 3 Ultra 550B A55B

via nebius

What changed

First observed in the model catalog

removed
GPT-OSS-120B

via berget

What changed

Removed from the model catalog

capability
MiMo-V2.5

via requesty

What changed

  • reasoningNoYesEnabled
context
GPT-5.6 Terra

via cortecs

What changed

  • Output limit1.05M128K−88%
removed
Nemotron-3-Nano-30B-A3B

via nebius

What changed

Removed from the model catalog

context
MiniMax-M2.7

via cortecs

What changed

  • Output limit196K197K1×
removed
MiniMax-M2.5-fast

via nebius

What changed

Removed from the model catalog

removed
Kimi-K2.5

via nebius

What changed

Removed from the model catalog

new model
Qwen3.8 Flash Next

via requesty

What changed

First observed in the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.1%
  • Output price$0.753/M$0.753/M▼ 0.1%
context
nvidia-nemotron-3-nano-omni

via cortecs

What changed

  • Output limit300K66K−78%
context
Hy3

via tencent-token-plan

What changed

  • Output limit64K128K2×
  • Input limit—192K
repriced
GLM Latest

via openrouter

What changed

  • Output limit944K236K−75%
  • Input price$1.17/M$1.15/M▼ 1.7%
  • Output price$3.96/M$3.50/M▼ 11.6%
  • Cache read$0.234/M$0.1/M▼ 57.3%
context
GPT-4o

via cortecs

What changed

  • Output limit128K16K−88%
context
Qwen3 235B-A22B Instruct 2507

via cortecs

What changed

  • Output limit131K262K2×
repriced
Kimi K2.5

via hyper

What changed

  • Input price$0.55/M$0.528/M▼ 4.0%
  • Output price$2.88/M$2.79/M▼ 3.5%
  • cache write$0.275/M$0.264/M▼ 4.0%
repriced
GLM-5.3-Flash

via requesty

What changed

  • Context1.05M1M−5%
  • Output limit131K262K2×
  • Input price$0.075/M$0.2/M▲ 166.7%
  • Output price$0.25/M$0.5/M▲ 100.0%
  • +1 more changes
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit131K393K3×
  • Input price$0.05/M$0.05/M▲ 0.0%
  • Output price$0.1/M$0.16/M▲ 60.1%
  • Cache read$0.0100/M$0.013/M▲ 30.1%
new model
DeepSeek V4 Flash 0731

via edenai

What changed

First observed in the model catalog

new model
Agentic Chat (Claude Fable 5.1)

via gitlab

What changed

First observed in the model catalog

context
Claude Haiku 4.5 (latest)

via cortecs

What changed

  • Output limit200K64K−68%
removed
GLM 4.7

via berget

What changed

Removed from the model catalog

new model
GLM Flash Latest

via openrouter

What changed

First observed in the model catalog

removed
Ornith 1.5 9B

via nano-gpt

What changed

Removed from the model catalog

context
Hy3

via edenai

What changed

  • Output limit64K128K2×
  • Input limit—192K
removed
Qwen3-32B

via nebius

What changed

Removed from the model catalog

context
Mixtral 8x7B Instruct v0.1

via cortecs

What changed

  • Output limit32K4K−87%
removed
Llama-3.3-70B-Instruct

via nebius

What changed

Removed from the model catalog

removed
KAT Coder Pro V2

via nano-gpt

What changed

Removed from the model catalog

removed
Ornith 1.5 9B Thinking

via nano-gpt

What changed

Removed from the model catalog

context
Nano Banana Pro

via kilo

What changed

  • Context131K66K−50%
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$1.03/M$1.02/M▼ 0.3%
  • Output price$2.05/M$2.05/M▼ 0.3%
  • Cache read$0.086/M$0.085/M▼ 0.3%
removed
GPT-4o Search Preview

via nano-gpt

What changed

Removed from the model catalog

removed
GLM-5

via nebius

What changed

Removed from the model catalog

context
GLM-5.3-Flash

via berget

What changed

  • last updated2026-08-292026-09-01
  • Context328K524K1.6×
context
Gemini 3.1 Flash Lite

via cortecs

What changed

  • Output limit1.05M66K−94%
new model
Nemotron 3.5 Lightning 30B A3B

via nebius

What changed

First observed in the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Cache read$0.15/M$0.015/M▼ 90.0%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.072/M$0.07/M▼ 2.1%
  • Output price$0.143/M$0.14/M▼ 2.1%
  • Cache read$0.014/M$0.014/M▼ 2.1%
removed
Hermes-4-70B

via nebius

What changed

Removed from the model catalog

context
nova-2-lite

via cortecs

What changed

  • Output limit1M66K−93%
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context1.05M262K−75%
  • Output limit944K236K−75%
  • Input price$1.17/M$1.15/M▼ 1.7%
  • Output price$3.96/M$3.50/M▼ 11.6%
  • +1 more changes
context
gemma-3-27b-it

via cortecs

What changed

  • Output limit131K110K−16%
removed
Cohere Command A+ (05/2026)

via nano-gpt

What changed

Removed from the model catalog

repriced
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit262K131K−50%
  • Cache read$0.25/M$0.2/M▼ 20.0%
deprecated
GPT-5.3 Chat (latest)

via openai

What changed

  • status—deprecated
context
Gemini 3.6 Flash

via cortecs

What changed

  • Output limit1.05M66K−94%
repriced
GPT OSS 20B (EU)

via edenai

What changed

  • Cache read$0.07/M$0.0070/M▼ 90.0%
new model
Muse Spark 1.3

via nano-gpt

What changed

First observed in the model catalog

removed
KAT Coder Air V2.5

via nano-gpt

What changed

Removed from the model catalog

context
MiniMax-M2.5

via cortecs

What changed

  • Output limit197K196K−0%
repriced
Nemotron 3 Ultra 550B A55B

via openrouter

What changed

  • Output limit16K33K2×
  • Input price$0.5/M$0.625/M▲ 25.0%
  • Output price$2.20/M$3.13/M▲ 42.0%
  • Cache read$0.1/M$0.188/M▲ 87.5%
context
GPT-5.6 Luna

via cortecs

What changed

  • Output limit1.05M128K−88%
new model
Gemini 3.8 Flash

via merge-gateway

What changed

First observed in the model catalog

context
GLM-5.3-Flash

via cortecs

What changed

  • Output limit64K1.05M16.4×
context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
removed
Muse Spark 1.3 Free

via opencode

What changed

Removed from the model catalog

context
Hy3

via huggingface

What changed

  • Output limit64K128K2×
  • Input limit—192K
removed
Llama-3.1-Nemotron-Ultra-253B-v1

via nebius

What changed

Removed from the model catalog

new model
GLM-5.3-Flash (EU)

via requesty

What changed

First observed in the model catalog

context
GPT-5.4

via cortecs

What changed

  • Output limit1.05M128K−88%
repriced
DeepSeek V4 Flash 0731

via requesty

What changed

  • reasoningNoYesEnabled
  • Output limit384K131K−66%
  • Input price$0.076/M$0.14/M▲ 84.2%
  • Output price$0.153/M$0.28/M▲ 83.0%
  • +1 more changes
context
Qwen3-Embedding-8B

via nebius

What changed

  • Context33K41K1.3×
  • Input limit33K41K1.3×
removed
Qwen3-235B-A22B-Thinking-2507-fast

via nebius

What changed

Removed from the model catalog

new model
Qwen3.8 2.4T A95B (EU)

via requesty

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new model
GLM 5.3 (50% off)

via vercel

What changed

First observed in the model catalog

new model
Gemini 3.8 Flash

via llmgateway

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via fireworks-ai

What changed

First observed in the model catalog

new model
Qwen3.8 Max 0902

via vercel

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash 0731

via nebius

What changed

First observed in the model catalog

context
mistral-7b-instruct-v0.2

via cortecs

What changed

  • Output limit32K8K−74%
removed
Qwen3-Next-80B-A3B-Thinking-fast

via nebius

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
GPT-5.6 Terra

via neon

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$12/M▼ 20.0%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write—$2.50/M
  • +1 more changes
context
Apertus 70B

via cortecs

What changed

  • Output limit66K16K−75%
removed
Hunyuan MT 7B

via nano-gpt

What changed

Removed from the model catalog

removed
DeepSeek Prover v2 671B

via nano-gpt

What changed

Removed from the model catalog

context
Hy3 Free

via opencode

What changed

  • Input limit—192K
new model
Gemini 3.8 Flash

via venice

What changed

First observed in the model catalog

repriced
GPT-5.6 Luna

via neon

What changed

  • Input price$1/M$0.2/M▼ 80.0%
  • Output price$6/M$1.20/M▼ 80.0%
  • Cache read$0.1/M$0.02/M▼ 80.0%
  • cache write—$0.25/M
  • +1 more changes
context
Claude Opus 4.5 (latest)

via cortecs

What changed

  • Output limit200K64K−68%
new model
Muse Spark 1.3

via openrouter

What changed

First observed in the model catalog

repriced
GLM-5.1

via hyper

What changed

  • Input price$1.33/M$1.29/M▼ 3.2%
  • Output price$4.31/M$4.22/M▼ 2.1%
  • cache write$0.666/M$0.645/M▼ 3.2%
repriced
Gemini Flash Latest

via nano-gpt

What changed

  • Input price$0.375/M$0.75/M▲ 100.0%
  • Output price$1.88/M$3.75/M▲ 100.0%
  • Cache read$0.037/M$0.075/M▲ 100.0%
  • cache write$0.021/M$0.075/M▲ 260.0%
context
MiniMax-M2

via cortecs

What changed

  • Output limit400K196K−51%
new model
Gemini 3.8 Flash

via openrouter

What changed

First observed in the model catalog

removed
GPT-4o mini Search Preview

via nano-gpt

What changed

Removed from the model catalog

context
Gemini 3.5 Flash Lite

via cortecs

What changed

  • Output limit1.05M66K−94%
new model
Muse Spark 1.3 Contributor

via openrouter

What changed

First observed in the model catalog

removed
DeepSeek-V3.2-fast

via nebius

What changed

Removed from the model catalog

removed
DeepSeek V4 Flash

via nebius

What changed

Removed from the model catalog

repriced
Claude Fable 5.1

via edenai

What changed

  • nameClaude Fable 5Claude Fable 5.1
  • descriptionClaude model for creative writing, analysis, and controlled agent workflowsClaude model for demanding reasoning and long-horizon agentic work
  • knowledge2026-01-312026-06
  • release date2026-06-092026-09-01
  • +2 more changes
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.753/M$0.753/M▼ 0.1%
  • Output price$0.753/M$0.753/M▼ 0.1%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.04/M$1.04/M▼ 0.1%
  • Output price$1.04/M$1.04/M▼ 0.1%
context
Hy3

via crossmodel

What changed

  • Input limit—192K
context
DeepSeek V4 Pro 0813

via cortecs

What changed

  • Output limit1.05M64K−94%
context
DeepSeek: DeepSeek V3.1

via kilo

What changed

  • Context161K164K1×
  • Output limit145K33K−77%
repriced
GPT-5.6 Sol

via azure

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
  • +1 more changes
removed
Sarvam 30B

via nano-gpt

What changed

Removed from the model catalog

context
GLM-5.2

via kilo

What changed

  • Output limit262K131K−50%
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit131K393K3×
  • Input price$0.05/M$0.05/M▲ 0.0%
  • Output price$0.1/M$0.16/M▲ 60.1%
  • Cache read$0.0100/M$0.013/M▲ 30.1%
removed
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

Removed from the model catalog

context
gpt-oss-120b

via nebius

What changed

  • Context128K131K1×
context
Qwen3-30B-A3B-Instruct-2507

via nebius

What changed

  • Context128K262K2×
  • Input limit120K262K2.2×
removed
Qwen3-Next-80B-A3B-Thinking

via nebius

What changed

Removed from the model catalog

context
Nemotron 3 Ultra 550B A55B

via kilo

What changed

  • Context262K256K−2%
  • Output limit16K33K2×
new model
Muse Spark 1.3 Contributor

via opencode-go

What changed

First observed in the model catalog

repriced
Gemini 3.8 Flash

via openrouter

What changed

  • familygeminigemini-flash
  • modalities.inputtext,image,video,pdf,audiotext,image,video,audio,pdf
  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • +3 more changes
new model
GLM-5.3 (Runware)

via llmgateway-providers

What changed

First observed in the model catalog

context
Hy3

via opencode-go

What changed

  • Output limit64K128K2×
  • Input limit—192K
context
Gemini 3.5 Flash

via cortecs

What changed

  • Output limit1.05M66K−94%
new model
Claude Fable 5.1

via opencode

What changed

First observed in the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
new model
Claude Fable 5.1

via crossmodel

What changed

First observed in the model catalog

new model
Claude Fable 5.1 (EU)

via requesty

What changed

First observed in the model catalog

context
Hy3 (Free)

via kenari

What changed

  • Output limit64K128K2×
  • Input limit—192K
context
GPT OSS 120B

via cortecs

What changed

  • Output limit128K131K1×
capability
Muse Spark 1.3 Free

via opencode

What changed

  • statusdeprecated—
context
GPT-4.1 mini

via cortecs

What changed

  • Output limit1.05M33K−97%
context
Hermes-4-405B

via nebius

What changed

  • Context128K131K1×
context
Claude Sonnet 4.5 (latest)

via cortecs

What changed

  • Output limit200K64K−68%
new model
Claude Fable 5.1

via requesty

What changed

First observed in the model catalog

new model
Qwen3.8-27B

via ovhcloud

What changed

First observed in the model catalog

new model
Muse Spark 1.3

via vercel

What changed

First observed in the model catalog

removed
MiniMax-M2.5

via nebius

What changed

Removed from the model catalog

removed
gpt-oss-120b-fast

via nebius

What changed

Removed from the model catalog

capability
Kimi K2.7 Code

via requesty

What changed

  • structured outputYesNoRemoved
removed
Kimi K2.6

via berget

What changed

Removed from the model catalog

new model
Gemini 3.8 Flash

via requesty

What changed

First observed in the model catalog

repriced
Gemini 3.8 Flash

via nano-gpt

What changed

  • descriptionGoogle's fast multimodal model for agentic workloads, including coding, tool use, image understanding, PDF and document extraction, audio, and video. Its capabilities, limits, reasoning behavior, and provisional pricing currently mirror Gemini 3.7 Flash.Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows
  • familygeminigemini-flash
  • Input price$0.375/M$0.75/M▲ 100.0%
  • Output price$1.88/M$3.75/M▲ 100.0%
  • +2 more changes
context
Hy3

via kilo

What changed

  • Input limit—192K
removed
GLM 4.6V Flash

via nano-gpt

What changed

Removed from the model catalog

context
Nova Pro 1.0

via cortecs

What changed

  • Output limit5K10K2×
new model
Claude Fable 5.1

via edenai

What changed

First observed in the model catalog

new model
GPT OSS 20B (Consensus Protocol)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Gemini 3.8 Flash

via opencode

What changed

First observed in the model catalog

removed
GPT-4 Turbo Preview

via nano-gpt

What changed

Removed from the model catalog

context
Hy3 (NovitaAI)

via llmgateway-providers

What changed

  • Input limit—192K
capability
MiMo-V2.5-Pro

via requesty

What changed

  • reasoningNoYesEnabled
context
mistral-nemo-instruct-2407

via cortecs

What changed

  • Output limit131K128K−2%
repriced
GLM-5.2

via openrouter

What changed

  • Output limit262K131K−50%
  • Input price$1.19/M$0.966/M▼ 18.8%
  • Output price$3.74/M$3.04/M▼ 18.8%
  • Cache read$0.221/M$0.193/M▼ 12.6%
context
nova-micro-v1

via cortecs

What changed

  • Output limit128K10K−92%
new model
Google: Gemini 3.8 Flash

via kilo

What changed

First observed in the model catalog

new model
Qwen3.8 Flash

via requesty

What changed

First observed in the model catalog

context
Hy3

via orcarouter

What changed

  • Output limit64K128K2×
  • Input limit—192K
context
Claude Opus 4.8

via cortecs

What changed

  • Output limit1M128K−87%
new model
Meta: Muse Spark 1.3

via kilo

What changed

First observed in the model catalog

new model
Grok 4.6

via azure

What changed

First observed in the model catalog

repriced
Gemini 3.7 Flash

via nano-gpt

What changed

  • Input price$0.375/M$0.75/M▲ 100.0%
  • Output price$1.88/M$3.75/M▲ 100.0%
  • Cache read$0.037/M$0.075/M▲ 100.0%
  • cache write$0.021/M$0.075/M▲ 260.0%
new model
Gemini 3.8 Flash

via google

What changed

First observed in the model catalog

removed
Llama 3.3 70B Instruct

via berget

What changed

Removed from the model catalog

capability
Fugu Ultra

via requesty

What changed

  • reasoningNoYesEnabled
repriced
GPT OSS 120B

via neon

What changed

  • Input price$0.072/M$0.15/M▲ 108.3%
  • Output price$0.28/M$0.6/M▲ 114.3%
removed
Qwen3.5-397B-A17B-fast

via nebius

What changed

Removed from the model catalog

deprecated
GPT-5.2 Chat

via openai

What changed

  • status—deprecated
removed
OpenAI o4-mini Deep Research

via nano-gpt

What changed

Removed from the model catalog

new model
Muse Spark 1.3 Contributor

via nano-gpt

What changed

First observed in the model catalog

context
Hy3

via kenari

What changed

  • Output limit64K128K2×
  • Input limit—192K
context
GLM-5.2

via cortecs

What changed

  • Output limit1.05M1M−5%
repriced
Qwen3Guard-Gen-8B

via ovhcloud

What changed

  • Input price—$0/M
  • Output price—$0/M
context
nova-lite-v1

via cortecs

What changed

  • Output limit300K10K−97%
context
Claude Sonnet 4 (latest)

via cortecs

What changed

  • Output limit200K65K−68%
context
GPT-4.1 nano

via cortecs

What changed

  • Output limit1.05M33K−97%
new model
Gemini 3.8 Flash

via google-vertex

What changed

First observed in the model catalog

removed
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

Removed from the model catalog

repriced
GPT OSS 20B

via edenai

What changed

  • Cache read$0.07/M$0.0070/M▼ 90.0%
new model
Gemini 3.8 Flash

via nano-gpt

What changed

First observed in the model catalog

removed
OpenAI o3 Deep Research

via nano-gpt

What changed

Removed from the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.1%
  • Output price$0.695/M$0.695/M▼ 0.1%
repriced
DeepSeek V4 Flash

via requesty

What changed

  • reasoningNoYesEnabled
  • structured outputNoYesEnabled
  • Context1M1.05M1×
  • Output limit384K131K−66%
  • +3 more changes
removed
Baichuan 4 Air

via nano-gpt

What changed

Removed from the model catalog

context
Kimi K2.6

via cortecs

What changed

  • Output limit256K262K1×
repriced
DeepSeek V3.1

via openrouter

What changed

  • Output limit145K33K−77%
  • Input price$0.55/M$0.25/M▼ 54.5%
  • Output price$1.65/M$0.95/M▼ 42.4%
  • Cache read$0.55/M$0.13/M▼ 76.4%
removed
OpenAI o1-preview

via nano-gpt

What changed

Removed from the model catalog

repriced
GLM-5.3

via llmgateway

What changed

  • Input price$1.30/M$1.20/M▼ 7.7%
  • Cache read$0.25/M$0.2/M▼ 20.0%
capability
Muse Spark 1.3 Contributor

via opencode-go

What changed

  • statusdeprecated—
context
Claude Sonnet 5

via cortecs

What changed

  • Output limit1M128K−87%
context
GPT-5 Nano

via cortecs

What changed

  • Output limit400K128K−68%
repriced
GPT OSS 20B

via neon

What changed

  • Input price$0.05/M$0.07/M▲ 40.0%
  • Output price$0.2/M$0.3/M▲ 50.0%
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$0.66/M$1.12/M▲ 69.0%
  • Output price$1.98/M$3.35/M▲ 69.0%
  • Cache read$0.022/M$0.037/M▲ 69.0%
removed
GLM-5.3

via berget

What changed

Removed from the model catalog

new model
Qwen3.8 Max 0902

via nano-gpt

What changed

First observed in the model catalog

context
Hy3 (free)

via orcarouter

What changed

  • Output limit64K128K2×
  • Input limit—192K
removed
QwenLong L1 32B

via nano-gpt

What changed

Removed from the model catalog

context
GPT-5.1

via cortecs

What changed

  • Output limit400K128K−68%
context
GPT-5 Mini

via cortecs

What changed

  • Output limit400K128K−68%
context
Hy3 (DeepInfra)

via llmgateway-providers

What changed

  • Input limit—192K
removed
KAT Coder Pro V2.5

via nano-gpt

What changed

Removed from the model catalog

removed
Kimi-K2.5-fast

via nebius

What changed

Removed from the model catalog

context
pixtral-12b-2409

via cortecs

What changed

  • Output limit128K4K−97%
context
GLM-5.2

via nebius

What changed

  • Context432K1.05M2.4×
  • Output limit432K1.05M2.4×
removed
GPT-5 Codex

via nano-gpt

What changed

Removed from the model catalog

repriced
Gemini 3.8 Flash

via kilo

What changed

  • nameGoogle: Gemini 3.8 FlashGemini 3.8 Flash
  • familygeminigemini-flash
  • modalities.inputtext,image,video,pdf,audiotext,image,video,audio,pdf
  • Input price$1.50/M$0.75/M▼ 50.0%
  • +4 more changes
context
Devstral 2

via cortecs

What changed

  • Output limit262K256K−2%
repriced
GPT-5.6 Sol

via neon

What changed

  • cache write—$6.25/M
  • tiers[object Object][object Object]
context
Gemini 3.7 Flash

via cortecs

What changed

  • Output limit1.05M66K−94%
context
Qwen3.6 35B-A3B

via cortecs

What changed

  • Output limit262K33K−87%
context
Hy3

via jalapeno

What changed

  • Output limit64K128K2×
  • Input limit—192K

103 events
removed
Claude Opus 5

via kilo

What changed

Removed from the model catalog

new model
Claude Fable 5.1

via llmgateway

What changed

First observed in the model catalog

capability
Nex AGI: Nex-N2-Mini (retires Sep 4)

via kilo

What changed

  • nameNex AGI: Nex-N2-MiniNex AGI: Nex-N2-Mini (retires Sep 4)
new model
Ornith 1.5 35B A3B

via iteracompute

What changed

First observed in the model catalog

new model
Claude Fable 5.1

via anthropic

What changed

First observed in the model catalog

repriced
Gemini 3.7 Flash

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.1%
  • Output price$0.696/M$0.695/M▼ 0.1%
new model
Claude Fable 5.1

via amazon-bedrock

What changed

First observed in the model catalog

new model
MiMo V2.5 Pro UltraSpeed

via vercel

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit393K131K−67%
  • Input price$0.05/M$0.05/M▼ 0.0%
  • Output price$0.16/M$0.1/M▼ 37.5%
  • Cache read$0.013/M$0.0100/M▼ 23.1%
repriced
MiniMax: MiniMax M3 (free)

via kilo

What changed

  • modalities.inputtext,imagetext,image,video
  • Output limit1.05M944K−10%
  • Cache read$0/M—
  • reasoning$0/M—
new model
Anthropic Claude Fable 5.1

via digitalocean

What changed

First observed in the model catalog

repriced
Magnum v4 72B

via openrouter

What changed

  • Input price$3/M$2.50/M▼ 16.7%
context
Qwen3.8 2.4T A95B

via kilo

What changed

  • Context1M262K−74%
  • Output limit262K131K−50%
removed
Claude Opus 4.7 (Fast)

via openrouter

What changed

Removed from the model catalog

repriced
Mancer: Weaver (alpha)

via kilo

What changed

  • Input price$0.5/M$0.4/M▼ 20.0%
context
DeepSeek V4 Pro

via kilo

What changed

  • Context1.02M1.05M1×
  • Output limit384K393K1×
repriced
MiniMax: MiniMax M2.7 (free)

via kilo

What changed

  • Output limit197K177K−10%
  • Cache read$0/M—
  • reasoning$0/M—
capability
Claude Fable 5.1

via kilo

What changed

  • nameAnthropic: Claude Fable 5.1 ($$$$)Claude Fable 5.1
  • knowledge—2026-06
repriced
Gemini 3.7 Flash (EU)

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
new model
Claude Fable 5.1

via merge-gateway

What changed

First observed in the model catalog

capability
Nex AGI: Nex-N2-Pro (retires Sep 4)

via kilo

What changed

  • nameNex AGI: Nex-N2-ProNex AGI: Nex-N2-Pro (retires Sep 4)
repriced
Devstral 2

via cortecs

What changed

  • Context262K256K−2%
  • Input price$0.446/M$0.478/M▲ 7.2%
  • Output price$2.23/M$2.39/M▲ 7.4%
removed
Claude Opus 4.8 (Fast)

via openrouter

What changed

Removed from the model catalog

context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.05M1.02M−2%
removed
Claude Fable 5.1

via opencode

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.464/M$0.464/M▼ 0.1%
  • Output price$0.928/M$0.927/M▼ 0.1%
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.424/M$0.404/M▼ 4.7%
  • Output price$1.61/M$1.50/M▼ 7.2%
  • cache write$0.212/M$0.202/M▼ 4.7%
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
repriced
Nano Banana 2

via requesty

What changed

  • tiers[object Object]
new model
Claude Fable 5.1

via vercel

What changed

First observed in the model catalog

removed
DeepSeek V3 0324

via vercel

What changed

Removed from the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Output limit384K393K1×
  • Input price$0.87/M$1.60/M▲ 83.9%
  • Output price$1.74/M$3.20/M▲ 83.9%
  • Cache read$0.072/M$0.135/M▲ 86.2%
repriced
Gemini 3.7 Flash

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.174/M$0.174/M▼ 0.1%
  • Output price$0.754/M$0.753/M▼ 0.1%
new model
Claude Fable 5.1 (Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Qwen3.8 27B TEE

via chutes

What changed

  • Input price$0.35/M$0.32/M▼ 8.6%
  • Output price$2.75/M$2.50/M▼ 9.1%
  • Cache read$0.035/M$0.032/M▼ 8.6%
capability
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • structured outputNoYesEnabled
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.754/M$0.753/M▼ 0.1%
  • Output price$0.754/M$0.753/M▼ 0.1%
removed
mistral-medium-2508

via cortecs

What changed

Removed from the model catalog

capability
Claude Fable 5.1

via vercel

What changed

  • temperatureYesNoRemoved
  • knowledge—2026-06
  • release date2026-08-312026-09-01
  • last updated2026-08-312026-09-01
repriced
Kimi K2.5

via hyper

What changed

  • Input price$0.544/M$0.55/M▲ 1.1%
  • Output price$2.85/M$2.88/M▲ 1.1%
  • cache write$0.272/M$0.275/M▲ 1.1%
repriced
ReMM SLERP 13B

via kilo

What changed

  • Input price$0.45/M$0.35/M▼ 22.2%
new model
Claude Fable 5.1

via opencode

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$0.66/M$1.12/M▲ 69.0%
  • Output price$1.98/M$3.35/M▲ 69.0%
  • Cache read$0.022/M$0.037/M▲ 69.0%
new model
Mercury 2.5 Preview

via nano-gpt

What changed

First observed in the model catalog

new model
Mercury 2.5 Preview

via openrouter

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.04/M$1.04/M▼ 0.1%
  • Output price$1.04/M$1.04/M▼ 0.1%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.08/M$0.03/M▼ 62.5%
  • Output price$0.18/M$0.1/M▼ 44.4%
new model
Claude Fable 5.1

via nano-gpt

What changed

First observed in the model catalog

new model
Claude Fable 5.1 (Global)

via amazon-bedrock

What changed

First observed in the model catalog

new model
Claude Fable 5.1 (US)

via amazon-bedrock

What changed

First observed in the model catalog

repriced
Claude Fable 5.1

via venice

What changed

  • Input price$10/M$12/M▲ 20.0%
  • Output price$50/M$60/M▲ 20.0%
  • Cache read$0.25/M$0.3/M▲ 20.0%
  • cache write$12.50/M$15/M▲ 20.0%
new model
Anthropic: Claude Fable 5.1 ($$$$)

via kilo

What changed

First observed in the model catalog

capability
Claude Fable 5.1

via openrouter

What changed

  • knowledge—2026-06
repriced
Gemini 3.1 Flash Lite (EU)

via requesty

What changed

  • tiers[object Object]
repriced
Magnum v4 72B

via kilo

What changed

  • Input price$3/M$2.50/M▼ 16.7%
repriced
Llama 4 Scout

via openrouter

What changed

  • Output limit8K16K2×
  • Input price$0.11/M$0.1/M▼ 9.1%
  • Output price$0.34/M$0.3/M▼ 11.8%
  • Cache read$0.055/M—
capability
Claude Fable 5.1

via nano-gpt

What changed

  • descriptionClaude Fable 5.1 improves on Fable 5 across agentic coding, long-running workflows, front-end and visual code generation, finance, analysis, and knowledge work, with more concise plans and summaries. Anthropic retains prompts and outputs for 30 days; Zero Data Retention is not available.Claude model for demanding reasoning and long-horizon agentic work
  • temperatureYesNoRemoved
  • knowledge—2026-06
repriced
Anthropic: Claude Fable Latest ($$$$)

via kilo

What changed

  • Cache read$1/M$0.25/M▼ 75.0%
repriced
Nano Banana Pro

via requesty

What changed

  • tiers[object Object]
repriced
Qwen3 235B A22B Instruct 2507

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.087/M$0.09/M▲ 2.9%
  • Output price$0.35/M$0.55/M▲ 57.1%
  • Cache read$0.018/M—
repriced
DeepSeek V4 Flash 0731

via fireworks-ai

What changed

  • Input price$0.14/M$0.22/M▲ 57.1%
  • Output price$0.28/M$0.66/M▲ 135.7%
  • Cache read$0.028/M$0.0070/M▼ 75.0%
capability
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • structured outputNoYesEnabled
new model
Claude Fable 5.1

via openrouter

What changed

First observed in the model catalog

removed
Claude Opus 4.7

via kilo

What changed

Removed from the model catalog

new model
Gemini 2.5 Flash-Lite (EU)

via requesty

What changed

First observed in the model catalog

removed
LongCat-2.0

via vancine

What changed

Removed from the model catalog

context
DeepSeek V4 Flash Vision Exp

via vercel

What changed

  • Context1M1.05M1×
  • Output limit384K1.05M2.7×
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit393K131K−67%
  • Input price$0.05/M$0.05/M▼ 0.0%
  • Output price$0.16/M$0.1/M▼ 37.5%
  • Cache read$0.013/M$0.0100/M▼ 23.1%
repriced
Qwen3.8 27B

via iteracompute

What changed

  • modalities.inputtext,image,videotext,image
  • Context262K328K1.3×
  • Output limit33K66K2×
  • Input limit—262K
  • +3 more changes
new model
Claude Fable 5.1

via venice

What changed

First observed in the model catalog

removed
Claude Opus 4.8

via kilo

What changed

Removed from the model catalog

repriced
Gemini 3.7 Flash

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
new model
Gemini 2.5 Pro (EU)

via requesty

What changed

First observed in the model catalog

new model
Step 3.7 Flash

via edenai

What changed

First observed in the model catalog

new model
Claude Fable 5.1

via azure

What changed

First observed in the model catalog

new model
Claude Fable 5.1

via google-vertex

What changed

First observed in the model catalog

new model
Claude Fable 5.1

via azure-cognitive-services

What changed

First observed in the model catalog

repriced
Gemini 3.7 Flash (US)

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
removed
Claude Opus 5 (Fast)

via openrouter

What changed

Removed from the model catalog

context
Meta: Llama 4 Scout

via kilo

What changed

  • Context131K328K2.5×
  • Output limit8K16K2×
new model
Inception: Mercury 2.5 Preview

via kilo

What changed

First observed in the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
context
Qwen: Qwen3 235B A22B Instruct 2507

via kilo

What changed

  • Output limit236K16K−93%
capability
Claude Fable 5.1

via merge-gateway

What changed

  • tool callNoYesEnabled
  • structured outputNoYesEnabled
  • modalities.inputtext,imagetext,image,pdf
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.188/M$0.19/M▲ 1.1%
  • Output price$0.7/M$0.63/M▼ 10.0%
  • cache write$0.094/M$0.095/M▲ 1.1%
repriced
Claude Fable Latest

via openrouter

What changed

  • Cache read$1/M$0.25/M▼ 75.0%
repriced
Claude Sonnet 5

via venice

What changed

  • Input price$2/M$3/M▲ 50.0%
  • Output price$10/M$15/M▲ 50.0%
  • Cache read$0.2/M$0.3/M▲ 50.0%
  • cache write$2.50/M$3.75/M▲ 50.0%
repriced
Gemini 3.1 Pro Preview

via requesty

What changed

  • tiers[object Object][object Object]
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.079/M$0.081/M▲ 1.9%
  • Output price$0.159/M$0.162/M▲ 1.9%
  • Cache read$0.016/M$0.016/M▲ 1.9%
repriced
Gemini 3.7 Flash

via edenai

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +2 more changes
repriced
Weaver (alpha)

via openrouter

What changed

  • Input price$0.5/M$0.4/M▼ 20.0%
removed
MiMo-V2.5-Pro

via vancine

What changed

Removed from the model catalog

repriced
MythoMax 13B

via openrouter

What changed

  • Output limit4K7K2×
  • Input price$0.06/M$0.4/M▲ 566.7%
  • Output price$0.06/M$0.6/M▲ 900.0%
repriced
Gemini 3.1 Flash Lite

via requesty

What changed

  • tiers[object Object]
new model
Kimi K2.7 Code Highspeed

via edenai

What changed

First observed in the model catalog

context
MythoMax 13B

via kilo

What changed

  • Context4K8K2×
  • Output limit4K7K2×
new model
Claude Fable 5.1

via google-vertex-anthropic

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
context
MiniMax-M2.5

via cortecs

What changed

  • Context197K196K−0%
capability
GLM-5.3 Flash (SCX.ai)

via llmgateway-providers

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image,video,pdf

150 events
removed
Hermes 4 405B

via llmgateway

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731

via requesty

What changed

  • reasoningYesNoRemoved
  • Output limit131K384K2.9×
  • Input price$0.14/M$0.076/M▼ 45.7%
  • Output price$0.28/M$0.153/M▼ 45.4%
  • +1 more changes
new model
Grok 4.6

via aihubmix

What changed

First observed in the model catalog

removed
Qwen3 32B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

context
Meta: Llama 4 Maverick

via kilo

What changed

  • Context1.05M128K−88%
  • Output limit16K115K7×
repriced
GLM-5.2

via requesty

What changed

  • Input price$1.20/M$0.8/M▼ 33.3%
  • Output price$4.20/M$2.55/M▼ 39.3%
  • Cache read$0.26/M$0.16/M▼ 38.5%
repriced
GLM Latest

via openrouter

What changed

  • Output limit131K944K7.2×
  • Input price$1.19/M$1.17/M▼ 1.5%
  • Output price$4.18/M$3.96/M▼ 5.3%
  • Cache read$0.247/M$0.234/M▼ 5.3%
removed
Gemma 3 27B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
GLM-5

via hyper

What changed

  • Input price$0.9/M$0.91/M▲ 1.1%
  • Output price$2.80/M$2.93/M▲ 4.6%
  • cache write$0.45/M$0.455/M▲ 1.1%
capability
Trinity Large Thinking

via openrouter

What changed

  • structured outputYesNoRemoved
new model
GLM-5.3-Flash

via aihubmix

What changed

First observed in the model catalog

repriced
Gemini 3.5 Flash Lite

via kilo

What changed

  • Input price$0.3/M$0.15/M▼ 50.0%
  • Output price$2.50/M$1.25/M▼ 50.0%
  • Cache read$0.03/M$0.015/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +1 more changes
repriced
Gemini 3.5 Flash

via kilo

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$9/M$4.50/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +1 more changes
context
Qwen3.5 35B A3B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
repriced
Gemini 3.1 Flash Lite

via kilo

What changed

  • Input price$0.25/M$0.125/M▼ 50.0%
  • Output price$1.50/M$0.75/M▼ 50.0%
  • Cache read$0.025/M$0.013/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +1 more changes
repriced
Gemini 3 Flash Preview

via kilo

What changed

  • Input price$0.5/M$0.25/M▼ 50.0%
  • Output price$3/M$1.50/M▼ 50.0%
  • Cache read$0.05/M$0.025/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +1 more changes
removed
KAT-Coder-Air V2.5

via openrouter

What changed

Removed from the model catalog

capability
Grok Code Fast 1

via opencode

What changed

  • attachmentYesNoRemoved
new model
GLM-5.3

via hyper

What changed

First observed in the model catalog

removed
Nemotron 3 Nano Omni (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

removed
Qwen3 30B A3B Instruct 2507 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

removed
Cosmos 3 Super Reasoner (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Qwen3 32B

via llmgateway

What changed

  • Input price$0.1/M$0.36/M▲ 260.0%
  • Output price$0.3/M$0.87/M▲ 190.0%
removed
Kimi K2.7 Code (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

removed
Nemotron 3 Nano 30B

via llmgateway

What changed

Removed from the model catalog

context
DeepSeek V4 Pro

via kilo

What changed

  • Context1.02M1.05M1×
  • Output limit384K393K1×
removed
Qwen3 30B A3B Instruct (2507)

via llmgateway

What changed

Removed from the model catalog

repriced
DeepSeek Chat

via kilo

What changed

  • Input price$0.4/M$0.257/M▼ 35.6%
  • Output price$1.30/M$1.03/M▼ 20.9%
repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.11/M$0.12/M▲ 9.1%
  • Output price$0.408/M$0.42/M▲ 2.9%
  • cache write$0.055/M$0.06/M▲ 9.1%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.174/M▼ 0.4%
  • Output price$0.757/M$0.754/M▼ 0.4%
new model
DeepSeek V4 Flash Vision Exp

via vancine

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
removed
Nemotron 3 Nano Omni

via llmgateway

What changed

Removed from the model catalog

removed
Mistral Medium 3

via empiriolabs

What changed

Removed from the model catalog

capability
Qwen 3.6 27B

via venice

What changed

  • open weightsNoYesEnabled
removed
Gemma 3 27B

via llmgateway

What changed

Removed from the model catalog

removed
Gemini Robotics-ER 1.6 Preview

via google

What changed

Removed from the model catalog

removed
Hermes 4 405B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

context
KAT-Coder-Pro V2.5

via openrouter

What changed

  • Output limit80K236K2.9×
context
Kwaipilot: KAT-Coder-Pro V2

via kilo

What changed

  • Context256K262K1×
  • Output limit80K144K1.8×
capability
GLM-5.3 Flash (Z AI)

via llmgateway-providers

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image,video,pdf
repriced
Gemini 3.6 Flash

via kilo

What changed

  • Input price$0.75/M$0.375/M▼ 50.0%
  • Output price$3.75/M$1.88/M▼ 50.0%
  • Cache read$0.075/M$0.037/M▼ 50.0%
  • cache write$0.042/M$0.021/M▼ 50.0%
  • +1 more changes
repriced
Trinity Large Thinking

via kilo

What changed

  • structured outputYesNoRemoved
  • Input price$0.22/M$0.25/M▲ 13.6%
  • Output price$0.85/M$0.8/M▼ 5.9%
repriced
GLM-5.3

via deepinfra

What changed

  • Cache read$0.24/M$0.12/M▼ 50.0%
removed
Kimi K3 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

new model
GLM-5.3 Flash (Runware)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
GLM-5.3

via requesty

What changed

  • structured outputYesNoRemoved
  • Context1M1.05M1×
  • Output limit128K1.05M8.2×
  • Input price$1.40/M$1.20/M▼ 14.3%
  • +1 more changes
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.404/M$0.424/M▲ 5.0%
  • Output price$1.50/M$1.61/M▲ 7.8%
  • cache write$0.202/M$0.212/M▲ 5.0%
removed
Cosmos 3 Super Reasoner

via llmgateway

What changed

Removed from the model catalog

removed
Hermes 4 70B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
removed
GLM-5.2 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Qwen3.7 Max

via crossmodel

What changed

  • Input price$1.50/M$1.88/M▲ 25.0%
  • Output price$4.50/M$5.63/M▲ 25.0%
  • Cache read$0.3/M$0.375/M▲ 25.0%
  • cache write$1.88/M$2.35/M▲ 25.0%
removed
DeepSeek V4 Pro (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

removed
Qwen2.5 VL 72B Instruct (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Qwen3-Next 80B-A3B Instruct

via openrouter

What changed

  • Output limit16K236K14.4×
  • Input price$0.09/M$0.1/M▲ 11.1%
  • Cache read—$0.07/M
new model
Qwen 3.8 Flash Next

via vercel

What changed

First observed in the model catalog

removed
Llama 3.3 70B Instruct (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Gemini 3.7 Flash

via kilo

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +1 more changes
removed
GLM 5.2 FP4

via coralbricks

What changed

Removed from the model catalog

repriced
Llama 4 Maverick 17B Instruct

via hyper

What changed

  • Input price$0.284/M$0.274/M▼ 3.5%
  • Output price$0.934/M$0.899/M▼ 3.7%
  • cache write$0.142/M$0.137/M▼ 3.5%
new model
GLM-5.3-Flash

via hyper

What changed

First observed in the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
removed
GPT OSS 120B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

new model
DeepSeek V4 Flash 0731

via aihubmix

What changed

First observed in the model catalog

removed
MiniCPM-V 4.5

via llmgateway

What changed

Removed from the model catalog

context
Qwen3.8 2.4T A95B

via kilo

What changed

  • Context1M262K−74%
  • Output limit262K131K−50%
repriced
GLM-5.3-Flash

via requesty

What changed

  • structured outputNoYesEnabled
  • Context1M131K−87%
  • Output limit128K131K1×
  • Input price$0.15/M$0.075/M▼ 50.0%
  • +2 more changes
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.466/M$0.464/M▼ 0.4%
  • Output price$0.931/M$0.928/M▼ 0.4%
new model
LongCat-2.0

via vancine

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.079/M$0.09/M▲ 14.2%
  • Output price$0.157/M$0.18/M▲ 14.2%
  • Cache read$0.016/M$0.018/M▲ 14.2%
repriced
GPT-5.6 Luna

via merge-gateway

What changed

  • cache write—$0.25/M
removed
Nemotron 3 Nano 30B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

removed
MiniCPM-V 4.5 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

context
Hy4 preview

via crossmodel

What changed

  • Output limit1.05M66K−94%
removed
MiniMax M3 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

removed
MiniMax M2.5 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

new model
Granite 4.2 8B

via nano-gpt

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.04/M▼ 0.4%
  • Output price$1.05/M$1.04/M▼ 0.4%
new model
Granite 4.2 8B

via openrouter

What changed

First observed in the model catalog

capability
GLM 5.3

via venice

What changed

  • open weightsNoYesEnabled
removed
Nemotron 3 Ultra 550B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
GPT-5.6 Sol

via merge-gateway

What changed

  • cache write—$5/M
new model
Hy4 preview

via crossmodel

What changed

First observed in the model catalog

repriced
Nano Banana

via kilo

What changed

  • Input price$0.3/M$0.15/M▼ 50.0%
  • Output price$2.50/M$1.25/M▼ 50.0%
  • Cache read$0.03/M$0.015/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
repriced
Kimi K2.5

via openrouter

What changed

  • Input price$0.6/M$0.45/M▼ 25.0%
  • Output price$3/M$2.25/M▼ 25.0%
  • Cache read$0.1/M$0.07/M▼ 30.0%
repriced
Qwen3.7 Plus

via crossmodel

What changed

  • Input price$0.288/M$0.32/M▲ 11.1%
  • Output price$1.13/M$1.25/M▲ 11.1%
  • Cache read$0.029/M$0.032/M▲ 11.1%
  • cache write$0.36/M$0.4/M▲ 11.1%
  • +1 more changes
removed
Qwen3 235B A22B Instruct 2507 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

new model
Hy4 preview

via vancine

What changed

First observed in the model catalog

new model
Abliterated Model Large V2

via abliteration-ai

What changed

First observed in the model catalog

capability
Kimi K2.5

via venice

What changed

  • open weightsNoYesEnabled
repriced
GLM-5.3 (EU)

via requesty

What changed

  • structured outputYesNoRemoved
  • Input price$1.75/M$1.20/M▼ 31.4%
  • Output price$4.50/M$4.20/M▼ 6.7%
  • Cache read$0.44/M$0.26/M▼ 40.9%
removed
Qwen3 Next 80B A3B Thinking (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Llama-3.3-70B-Instruct

via llmgateway

What changed

  • Input price$0.13/M$0.135/M▲ 3.8%
removed
Kimi K2.6 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Gemini 3.1 Flash Lite Preview

via kilo

What changed

  • Input price$0.25/M$0.125/M▼ 50.0%
  • Output price$1.50/M$0.75/M▼ 50.0%
  • Cache read$0.025/M$0.013/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
  • +1 more changes
new model
GLM 5.3 TEE

via nano-gpt

What changed

First observed in the model catalog

repriced
Gemini 3.1 Pro Preview

via kilo

What changed

  • Input price$2/M$1/M▼ 50.0%
  • Output price$12/M$6/M▼ 50.0%
  • Cache read$0.2/M$0.1/M▼ 50.0%
  • cache write$0.375/M$0.188/M▼ 50.0%
  • +1 more changes
repriced
Claude Sonnet 5

via merge-gateway

What changed

  • Input price$2/M$3/M▲ 50.0%
  • Output price$10/M$15/M▲ 50.0%
repriced
Grok 4.6

via merge-gateway

What changed

  • Input price$1.50/M$2/M▲ 33.3%
  • Output price$4.50/M$6/M▲ 33.3%
  • Cache read$0.375/M$0.5/M▲ 33.3%
removed
Tencent: Hy3 (free)

via kilo

What changed

Removed from the model catalog

new model
Qwen3.8 27B

via groq

What changed

First observed in the model catalog

removed
Llama 3.1 Nemotron Ultra 253B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

new model
MiMo-V2.5-Pro

via vancine

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Input price$0.03/M$0.05/M▲ 66.7%
context
KAT-Coder-Pro V2

via openrouter

What changed

  • Output limit80K144K1.8×
removed
Nemotron 3 Super 120B

via llmgateway

What changed

Removed from the model catalog

removed
GLM-5.1 (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit262K131K−50%
  • Cache read$0.25/M$0.2/M▼ 20.0%
removed
Llama 3.1 Nemotron Ultra 253B

via llmgateway

What changed

Removed from the model catalog

repriced
Gemini 3.5 Flash

via openrouter

What changed

  • Input price$1.50/M$1.65/M▲ 10.0%
  • Output price$9/M$9.90/M▲ 10.0%
  • Cache read$0.15/M$0.165/M▲ 10.0%
  • reasoning$9/M$9.90/M▲ 10.0%
new model
DeepSeek V4 Pro 0813

via aihubmix

What changed

First observed in the model catalog

new model
GLM 5.3 FP4

via coralbricks

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.174/M▼ 0.4%
  • Output price$0.699/M$0.696/M▼ 0.4%
repriced
Nano Banana Pro

via kilo

What changed

  • Input price$2/M$1/M▼ 50.0%
  • Output price$12/M$6/M▼ 50.0%
  • Cache read$0.2/M$0.1/M▼ 50.0%
  • cache write$0.375/M$0.188/M▼ 50.0%
  • +1 more changes
new model
Qwen3.8 2.4T A95B

via aihubmix

What changed

First observed in the model catalog

context
Hy3

via crossmodel

What changed

  • Output limit262K131K−50%
capability
DeepSeek V4 Pro 0813

via venice

What changed

  • open weightsNoYesEnabled
new model
Abliterated Model Large V2

via nano-gpt

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
GPT-5.6 Terra

via merge-gateway

What changed

  • cache write—$2.50/M
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.31/M$1.33/M▲ 1.4%
  • Output price$4.27/M$4.31/M▲ 1.0%
  • cache write$0.657/M$0.666/M▲ 1.4%
removed
Nemotron 3 Super 120B (Nebius AI)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
GLM Latest

via nano-gpt

What changed

  • Input price$0.42/M$1/M▲ 138.1%
  • Output price$1.32/M$3.20/M▲ 142.4%
  • Cache read$0.078/M$0.2/M▲ 156.4%
capability
Kimi K2.5

via moonshotai-cn

What changed

  • attachmentNoYesEnabled
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.178/M$0.188/M▲ 5.6%
  • Output price$0.68/M$0.7/M▲ 2.9%
  • cache write$0.089/M$0.094/M▲ 5.6%
repriced
Gemma 4 12B Instruct

via nano-gpt

What changed

  • Input price$0.06/M$0.05/M▼ 16.7%
  • Output price$0.3/M$0.25/M▼ 16.7%
  • Cache read$0.03/M$0.025/M▼ 16.7%
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context262K1.05M4×
  • Output limit131K944K7.2×
  • Input price$1.19/M$1.17/M▼ 1.5%
  • Output price$4.18/M$3.96/M▼ 5.3%
  • +1 more changes
removed
Hermes 4 70B

via llmgateway

What changed

Removed from the model catalog

repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$0.66/M$1.32/M▲ 100.0%
  • Output price$1.98/M$3.96/M▲ 100.0%
  • Cache read$0.022/M$0.044/M▲ 100.0%
context
Qwen3-Next 80B-A3B Instruct

via kilo

What changed

  • Output limit16K236K14.4×
capability
Qwen3.8 27B

via llmtech

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
context
Kwaipilot: KAT-Coder-Pro V2.5

via kilo

What changed

  • Context256K262K1×
  • Output limit80K236K2.9×
new model
DeepSeek V4 Pro 0813

via cortecs

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Output limit384K393K1×
  • Input price$0.417/M$1.60/M▲ 283.5%
  • Output price$0.835/M$3.20/M▲ 283.5%
  • Cache read$0.035/M$0.135/M▲ 288.3%
repriced
DeepSeek: DeepSeek V3 0324

via kilo

What changed

  • Input price$0.27/M$0.25/M▼ 7.4%
  • Output price$1.12/M$1/M▼ 10.7%
  • Cache read$0.135/M—
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.757/M$0.754/M▼ 0.4%
  • Output price$0.757/M$0.754/M▼ 0.4%
repriced
Llama 4 Maverick

via openrouter

What changed

  • Output limit16K115K7×
  • Output price$0.8/M$0.696/M▼ 13.0%
removed
Kwaipilot: KAT-Coder-Air V2.5

via kilo

What changed

Removed from the model catalog

new model
Qwen3.7 Flash

via aihubmix

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash 0731

via merge-gateway

What changed

First observed in the model catalog

removed
Qwen2.5-VL 72B Instruct

via llmgateway

What changed

Removed from the model catalog

new model
GLM-5.3

via aihubmix

What changed

First observed in the model catalog

new model
IBM: Granite 4.2 8B

via kilo

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Input price$0.03/M$0.05/M▲ 66.7%
context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
repriced
Kimi K3

via merge-gateway

What changed

  • Input price$3/M$2.90/M▼ 3.3%
  • Output price$15/M$14/M▼ 6.7%
capability
Kimi K2.5

via moonshotai

What changed

  • attachmentNoYesEnabled

291 events
removed
GLM 4.5 (Thinking)

via nano-gpt

What changed

Removed from the model catalog

new model
Opus Distill

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3 (NovitaAI)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
repriced
GLM-5.1

via openrouter

What changed

  • Output limit182K128K−30%
  • Input price$1.26/M$0.966/M▼ 23.3%
  • Output price$3.96/M$3.04/M▼ 23.3%
  • Cache read$0.234/M$0.179/M▼ 23.3%
new model
GLM 4.6 Original

via nano-gpt

What changed

First observed in the model catalog

new model
GLM 5.3

via nano-gpt

What changed

First observed in the model catalog

capability
Qwen 3.8 27B Uncensored

via nano-gpt

What changed

  • release date2026-08-212026-07-29
capability
GLM-5.3

via cortecs

What changed

  • open weightsNoYesEnabled
repriced
Qwen3.8 27B

via nano-gpt

What changed

  • Input price$0.2/M$0.15/M▼ 25.0%
  • Output price$1.40/M$0.7/M▼ 50.0%
removed
Jamba Mini 1.7

via nano-gpt

What changed

Removed from the model catalog

new provider
DeepSeek V4 Flash Vision (Exp)

via above

What changed

above began listing this model

removed
GLM 4.7 Original

via nano-gpt

What changed

Removed from the model catalog

new provider
GLM-5.2

via sensenova

What changed

sensenova began listing this model

capability
Qwen 3.8 27B Obliterated Thinking

via nano-gpt

What changed

  • release date2026-08-242026-07-29
capability
Qwen 3.8 27B Fable

via nano-gpt

What changed

  • release date2026-08-282026-07-29
repriced
Gemma 4 26B A4B Uncensored Thinking

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
repriced
Fabled

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
repriced
Mistral Medium (latest)

via edenai

What changed

  • Cache read—$0.15/M
new provider
MiMo V2.5 Pro

via above

What changed

above began listing this model

new model
GLM 4.7 Flash Thinking

via nano-gpt

What changed

First observed in the model catalog

new model
GLM 4.6V Flash

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3

via cline-pass

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3

via edenai

What changed

  • open weightsNoYesEnabled
new provider
Kloker Integration Developer

via klokintegration

What changed

klokintegration began listing this model

new model
GLM-5.3-Flash

via berget

What changed

First observed in the model catalog

removed
Abliterated Model Large

via nano-gpt

What changed

Removed from the model catalog

capability
Gemma 4 26B A4B MeroMero Thinking

via nano-gpt

What changed

  • release date2026-08-262026-07-29
new model
GLM 5

via nano-gpt

What changed

First observed in the model catalog

repriced
Qwen3.8 27B Thinking

via nano-gpt

What changed

  • Input price$0.2/M$0.15/M▼ 25.0%
  • Output price$1.40/M$0.7/M▼ 50.0%
removed
Gembrain

via nano-gpt

What changed

Removed from the model catalog

repriced
Mistral Nemo

via kilo

What changed

  • Input price$0.165/M$0.019/M▼ 88.5%
  • Output price$0.165/M$0.03/M▼ 81.8%
  • Cache read$0.017/M—
repriced
Ornith 1.5 9B Thinking

via nano-gpt

What changed

  • Input price$0.05/M$0.1/M▲ 100.0%
  • Output price$0.1/M$0.2/M▲ 100.0%
  • Cache read$0.025/M$0.05/M▲ 100.0%
new model
GLM 5.3 Flash

via above

What changed

First observed in the model catalog

new model
GLM 5.2 Thinking

via nano-gpt

What changed

First observed in the model catalog

repriced
GPT-4.1 nano

via openrouter

What changed

  • Output limit943K33K−97%
  • Cache read$0.03/M$0.025/M▼ 16.7%
removed
GLM 4.5 Air (Thinking)

via nano-gpt

What changed

Removed from the model catalog

removed
End-to-End Encrypted

via trustedrouter

What changed

Removed from the model catalog

repriced
Mistral Medium 3.5

via edenai

What changed

  • Cache read—$0.15/M
repriced
Qwen3-Next 80B-A3B Instruct

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.1/M$0.09/M▼ 10.0%
  • Cache read$0.07/M—
repriced
DarkIdol

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
new model
GLM 4.7 Original Thinking

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 4.6 Turbo

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3

via zai

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3

via aiand

What changed

  • open weightsNoYesEnabled
new model
Synth

via trustedrouter

What changed

First observed in the model catalog

capability
GLM-5.3

via zhipuai

What changed

  • open weightsNoYesEnabled
repriced
Luminous Mirror

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
capability
GLM-5.3

via crof

What changed

  • open weightsNoYesEnabled
new provider
MiMo V2.5 Pro UltraSpeed

via above

What changed

above began listing this model

removed
Dark Soul

via nano-gpt

What changed

Removed from the model catalog

new model
MiniMax H3 Max

via vercel

What changed

First observed in the model catalog

removed
GLM 5 Thinking

via nano-gpt

What changed

Removed from the model catalog

repriced
Gembrain

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
removed
GLM 4.5 Air

via nano-gpt

What changed

Removed from the model catalog

capability
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

  • release date2026-08-212026-07-29
context
Glm 5.3 Flash

via cloudflare-workers-ai

What changed

  • Context1.31M1.05M−20%
  • Output limit1.31M1.05M−20%
removed
Step R1 V Mini

via nano-gpt

What changed

Removed from the model catalog

removed
Musica

via nano-gpt

What changed

Removed from the model catalog

new provider
Nemotron 3 Ultra (free)

via bothub

What changed

bothub began listing this model

capability
GLM-5.3

via zhipuai-coding-plan

What changed

  • open weightsNoYesEnabled
removed
Abliterated Model

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 5.1 Thinking

via nano-gpt

What changed

First observed in the model catalog

repriced
GLM 5.3 Flash Uncensored

via nano-gpt

What changed

  • Input price$0.125/M$0.35/M▲ 180.0%
  • Output price$0.5/M$1.40/M▲ 180.0%
  • Cache read$0.063/M$0.175/M▲ 180.0%
new model
Moonlight Dusk

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 4.7 Original Thinking

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3

via togetherai

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3

via requesty

What changed

  • open weightsNoYesEnabled
new model
DarkIdol

via nano-gpt

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via orcarouter

What changed

First observed in the model catalog

removed
Jamba Mini

via nano-gpt

What changed

Removed from the model catalog

removed
Synth Code

via trustedrouter

What changed

Removed from the model catalog

repriced
Moonlight Dusk

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
new provider
Qwen 3.8 Max

via above

What changed

above began listing this model

repriced
Chimera X

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
capability
Nvidia Nemotron 3.5 Lightning Thinking

via nano-gpt

What changed

  • structured outputYesNoRemoved
capability
GLM-5.3 (free)

via tokenrouter

What changed

  • open weightsNoYesEnabled
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.529/M$0.508/M▼ 3.8%
  • Output price$1.06/M$1.02/M▼ 3.8%
  • Cache read$0.044/M$0.042/M▼ 3.9%
capability
Gemma 4 31B MeroMero v2

via nano-gpt

What changed

  • release date2026-08-232026-07-29
removed
GLM Zero Preview

via nano-gpt

What changed

Removed from the model catalog

capability
GLM 5.3

via fireworks-ai

What changed

  • open weightsNoYesEnabled
repriced
Shadow Siren

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
removed
GLM 5 Original

via nano-gpt

What changed

Removed from the model catalog

removed
GLM 4.5

via nano-gpt

What changed

Removed from the model catalog

new model
Abliterated Model Large

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3

via deepinfra

What changed

  • open weightsNoYesEnabled
removed
GLM-4 AirX

via nano-gpt

What changed

Removed from the model catalog

repriced
Kimi Latest

via nano-gpt

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$13.50/M$10/M▼ 25.9%
  • Cache read$0.25/M$0.2/M▼ 20.0%
repriced
Dark Soul

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
repriced
DeepSeek V3.2

via openrouter

What changed

  • Output limit66K147K2.3×
  • Input price$0.269/M$0.26/M▼ 3.3%
  • Output price$0.4/M$0.38/M▼ 5.0%
  • Cache read$0.135/M$0.13/M▼ 3.3%
new model
GLM 5.3

via neuralwatt

What changed

First observed in the model catalog

repriced
Devstral 2

via edenai

What changed

  • Input price$0.44/M$0.4/M▼ 9.1%
  • Output price$2.20/M$2/M▼ 9.1%
  • Cache read$0.044/M$0.04/M▼ 9.1%
capability
GLM 5.3

via venice

What changed

  • open weightsNoYesEnabled
removed
GLM Latest

via nano-gpt

What changed

Removed from the model catalog

new provider
GLM 5.2 Fast

via above

What changed

above began listing this model

removed
Hunyuan Turbo S

via nano-gpt

What changed

Removed from the model catalog

new model
Shadow Siren

via nano-gpt

What changed

First observed in the model catalog

context
DeepSeek V3.2

via kilo

What changed

  • Output limit66K147K2.3×
new model
GLM-5.3-Flash

via synthetic

What changed

First observed in the model catalog

removed
Gemma 4 26B A4B Uncensored Thinking

via nano-gpt

What changed

Removed from the model catalog

new model
Novelist

via nano-gpt

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B MeroMero

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
capability
GLM 5.3 Thinking

via nano-gpt

What changed

  • open weightsNoYesEnabled
removed
Zero Data Retention

via trustedrouter

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output price$0.14/M$0.16/M▲ 14.3%
removed
Garnet

via nano-gpt

What changed

Removed from the model catalog

new provider
DeepSeek V4 Pro

via above

What changed

above began listing this model

capability
GLM 5.3

via nano-gpt

What changed

  • open weightsNoYesEnabled
context
Qwen3-Next 80B-A3B Instruct

via kilo

What changed

  • Output limit236K16K−93%
new provider
Gemma 4 31B IT (free)

via bothub

What changed

bothub began listing this model

new model
GLM 5.3 Thinking

via nano-gpt

What changed

First observed in the model catalog

capability
Glm 5.3

via cloudflare-workers-ai

What changed

  • open weightsNoYesEnabled
removed
DarkIdol

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 4.6V

via nano-gpt

What changed

First observed in the model catalog

new model
GLM 4.7 Flash Original

via nano-gpt

What changed

First observed in the model catalog

new model
GLM-5.3

via berget

What changed

First observed in the model catalog

capability
GLM-5.3

via openrouter

What changed

  • open weightsNoYesEnabled
repriced
Qwen 3.8 27B Fable

via nano-gpt

What changed

  • Input price$0.2/M$0.25/M▲ 25.0%
  • Output price$1.40/M$1.50/M▲ 7.1%
  • Cache read$0.04/M$0.125/M▲ 212.5%
new model
GLM 4.5 (Thinking)

via nano-gpt

What changed

First observed in the model catalog

new model
Gemsicle

via nano-gpt

What changed

First observed in the model catalog

removed
Jamba Large 1.6

via nano-gpt

What changed

Removed from the model catalog

removed
GLM 4.6V Original

via nano-gpt

What changed

Removed from the model catalog

removed
GLM 5.2

via nano-gpt

What changed

Removed from the model catalog

context
GLM-5.1

via kilo

What changed

  • Context203K200K−1%
  • Output limit182K128K−30%
repriced
Gemma 4 26B A4B

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
new model
End-to-End Encrypted

via trustedrouter

What changed

First observed in the model catalog

removed
Ornith 1.5 397B

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3 (Baidu)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
removed
GLM 4.6 Original

via nano-gpt

What changed

Removed from the model catalog

capability
GLM 5.3

via baseten

What changed

  • open weightsNoYesEnabled
new model
GLM 4.7

via nano-gpt

What changed

First observed in the model catalog

new model
Qwen3.8 27B

via aiand

What changed

First observed in the model catalog

capability
GLM5.3

via digitalocean

What changed

  • open weightsNoYesEnabled
capability
Nvidia Nemotron 3.5 Lightning

via nano-gpt

What changed

  • structured outputYesNoRemoved
repriced
Mistral: Mistral Small 3.2 24B

via kilo

What changed

  • Input price$0.11/M$0.075/M▼ 31.8%
  • Output price$0.33/M$0.2/M▼ 39.4%
  • Cache read$0.011/M—
capability
Qwen 3.8 27B Uncensored Thinking

via nano-gpt

What changed

  • release date2026-08-252026-07-29
removed
GLM 5.3 Preview Thinking

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 5.3 Flash

via fireworks-ai

What changed

First observed in the model catalog

repriced
Qwen 3.8 27B Obliterated

via nano-gpt

What changed

  • Input price$0.18/M$0.25/M▲ 38.9%
  • Output price$0.5/M$1.50/M▲ 200.0%
  • Cache read$0.075/M$0.125/M▲ 66.7%
capability
Gemma 4 31B MeroMero v2 Thinking

via nano-gpt

What changed

  • release date2026-08-242026-07-29
removed
Luminous Mirror

via nano-gpt

What changed

Removed from the model catalog

removed
Synth

via trustedrouter

What changed

Removed from the model catalog

new model
Auto

via trustedrouter

What changed

First observed in the model catalog

capability
GLM-5.3

via tokengo

What changed

  • open weightsNoYesEnabled
removed
Shadow Siren

via nano-gpt

What changed

Removed from the model catalog

capability
GLM 5.3

via vercel

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3 (Z AI)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
new model
Synth Code

via trustedrouter

What changed

First observed in the model catalog

context
GPT-4.1 nano

via kilo

What changed

  • Output limit943K33K−97%
removed
Jamba Mini 1.6

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 4.5

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3

via llmgateway

What changed

  • open weightsNoYesEnabled
removed
Ornith 1.5 397B Thinking

via nano-gpt

What changed

Removed from the model catalog

removed
Fast

via trustedrouter

What changed

Removed from the model catalog

repriced
Trinity Large Thinking

via openrouter

What changed

  • Output limit236K80K−66%
  • Input price$0.22/M$0.25/M▲ 13.6%
  • Output price$0.85/M$0.8/M▼ 5.9%
new model
Gemma 4 26B A4B Uncensored

via nano-gpt

What changed

First observed in the model catalog

capability
Ornith 1.5 9B

via nano-gpt

What changed

  • release date2026-08-232026-07-29
repriced
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

  • Output price$0.5/M$0.95/M▲ 90.0%
repriced
Gemma 4 26B A4B MeroMero Thinking

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
capability
GLM-5.3

via kenari

What changed

  • open weightsNoYesEnabled
removed
Cheap

via trustedrouter

What changed

Removed from the model catalog

removed
GLM-4 Air

via nano-gpt

What changed

Removed from the model catalog

new provider
GLM 5.2

via above

What changed

above began listing this model

capability
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

  • release date2026-08-242026-07-29
new model
GLM 4.7 Thinking

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 4.7 Thinking

via nano-gpt

What changed

Removed from the model catalog

new model
Luminous Mirror

via nano-gpt

What changed

First observed in the model catalog

removed
Claude 4 Opus Thinking (32K)

via nano-gpt

What changed

Removed from the model catalog

repriced
Muse Glimmer 30B

via edenai

What changed

  • Output price$1.20/M$1.10/M▼ 8.3%
repriced
Kimi K3

via nano-gpt

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$13.50/M$10/M▼ 25.9%
  • Cache read$0.25/M$0.2/M▼ 20.0%
capability
GLM-5.3

via vancine

What changed

  • open weightsNoYesEnabled
new model
GLM 4.5 Air (Thinking)

via nano-gpt

What changed

First observed in the model catalog

removed
Claude 4.1 Opus Thinking (32K)

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 4.6 Turbo (Thinking)

via nano-gpt

What changed

First observed in the model catalog

new model
Garnet

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 5

via nano-gpt

What changed

Removed from the model catalog

repriced
Ornith 1.5 9B

via nano-gpt

What changed

  • Input price$0.05/M$0.1/M▲ 100.0%
  • Output price$0.1/M$0.2/M▲ 100.0%
  • Cache read$0.025/M$0.05/M▲ 100.0%
new model
Fabled

via nano-gpt

What changed

First observed in the model catalog

new model
Cheap

via trustedrouter

What changed

First observed in the model catalog

repriced
Llama 4 Maverick

via openrouter

What changed

  • Output limit16K115K7×
  • Output price$0.8/M$0.696/M▼ 13.0%
removed
Gemini LearnLM Experimental

via nano-gpt

What changed

Removed from the model catalog

repriced
Devstral 2

via openrouter

What changed

  • Input price$0.44/M$0.4/M▼ 9.1%
  • Output price$2.20/M$2/M▼ 9.1%
  • Cache read$0.044/M$0.04/M▼ 9.1%
removed
GLM 4.7 Flash Original

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 4.7 Flash

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 5 Original Thinking

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 5 Thinking

via nano-gpt

What changed

First observed in the model catalog

new model
Zero Data Retention

via trustedrouter

What changed

First observed in the model catalog

new provider
Kloker

via klokintegration

What changed

klokintegration began listing this model

new model
Gembrain

via nano-gpt

What changed

First observed in the model catalog

removed
Fabled

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3

via crossmodel

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3

via kilo

What changed

  • open weightsNoYesEnabled
removed
Jamba Large 1.7

via nano-gpt

What changed

Removed from the model catalog

removed
GLM 4.6V Flash

via nano-gpt

What changed

Removed from the model catalog

removed
GLM-4 Plus

via nano-gpt

What changed

Removed from the model catalog

repriced
Qwen 3.8 27B Obliterated Thinking

via nano-gpt

What changed

  • Input price$0.18/M$0.25/M▲ 38.9%
  • Output price$0.5/M$1.50/M▲ 200.0%
  • Cache read$0.075/M$0.125/M▲ 66.7%
repriced
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

  • Output price$0.5/M$0.95/M▲ 90.0%
capability
GLM-5.3

via orcarouter

What changed

  • open weightsNoYesEnabled
new provider
DeepSeek V4 Flash

via sensenova

What changed

sensenova began listing this model

capability
GLM-5.3 Highspeed

via zhipuai-coding-plan

What changed

  • open weightsNoYesEnabled
removed
GLM 4.7

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 5.1

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3 (SCX.ai)

via llmgateway-providers

What changed

  • open weightsNoYesEnabled
repriced
Qwen 3.8 27B Uncensored Thinking

via nano-gpt

What changed

  • Input price$0.18/M$0.25/M▲ 38.9%
  • Output price$0.5/M$1.50/M▲ 200.0%
  • Cache read$0.075/M$0.125/M▲ 66.7%
removed
GLM 4.6 Turbo (Thinking)

via nano-gpt

What changed

Removed from the model catalog

new provider
SenseNova 6.8 Flash Lite

via sensenova

What changed

sensenova began listing this model

capability
GLM-5.3 Highspeed

via zai-coding-plan

What changed

  • open weightsNoYesEnabled
repriced
Isometry

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output price$0.1/M$0.14/M▲ 40.0%
  • Cache read$0.0070/M$0.01/M▲ 42.9%
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
new model
GLM 5 Original

via nano-gpt

What changed

First observed in the model catalog

capability
Ornith 1.5 35B

via nano-gpt

What changed

  • release date2026-08-202026-07-29
capability
GLM-5.3

via friendli

What changed

  • open weightsNoYesEnabled
capability
GLM-5.3

via volcengine-coding-plan

What changed

  • open weightsNoYesEnabled
removed
Chimera X

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3

via zai-coding-plan

What changed

  • open weightsNoYesEnabled
repriced
Qwen 3.8 27B Uncensored

via nano-gpt

What changed

  • Input price$0.18/M$0.25/M▲ 38.9%
  • Output price$0.5/M$1.50/M▲ 200.0%
  • Cache read$0.075/M$0.125/M▲ 66.7%
new model
Musica

via nano-gpt

What changed

First observed in the model catalog

capability
GLM 5.3

via empiriolabs

What changed

  • open weightsNoYesEnabled
new model
Fast

via trustedrouter

What changed

First observed in the model catalog

repriced
Gemma 4 31B

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
context
Meta: Llama 4 Maverick

via kilo

What changed

  • Context1.05M128K−88%
  • Output limit16K115K7×
new model
GLM 4.6 Turbo

via nano-gpt

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B Uncensored

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
repriced
Codestral (latest)

via edenai

What changed

  • Cache read—$0.03/M
new model
Isometry

via nano-gpt

What changed

First observed in the model catalog

context
Trinity Large Thinking

via kilo

What changed

  • Output limit236K80K−66%
new provider
Kloker Integration Architect

via klokintegration

What changed

klokintegration began listing this model

new model
GLM 4.7 Original

via nano-gpt

What changed

First observed in the model catalog

new model
GLM-5.3

via aiand

What changed

First observed in the model catalog

removed
GLM 5.3 Preview

via nano-gpt

What changed

Removed from the model catalog

repriced
Gemma 4 31B MeroMero v2

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.081/M$0.081/M▼ 0.7%
  • Output price$0.163/M$0.162/M▼ 0.7%
  • Cache read$0.016/M$0.016/M▼ 0.7%
new model
Chimera X

via nano-gpt

What changed

First observed in the model catalog

new model
GLM 4.7 Flash Original Thinking

via nano-gpt

What changed

First observed in the model catalog

removed
Moonlight Dusk

via nano-gpt

What changed

Removed from the model catalog

repriced
Garnet

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
removed
GLM 4.7 Flash Original Thinking

via nano-gpt

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
removed
GLM 4.7 Flash Thinking

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3 (EU)

via requesty

What changed

  • open weightsNoYesEnabled
repriced
Mistral Small (latest)

via edenai

What changed

  • Cache read—$0.015/M
removed
Isometry

via nano-gpt

What changed

Removed from the model catalog

new model
GLM Latest

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 5.1

via nano-gpt

What changed

Removed from the model catalog

new provider
DeepSeek V4 Flash

via above

What changed

above began listing this model

context
Llama 3.1 70B Dracarys 2

via nano-gpt

What changed

  • Context16K33K2×
  • Input limit16K33K2×
new provider
GLM-5.3 (free)

via tokenrouter

What changed

tokenrouter began listing this model

repriced
Opus Distill

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
removed
GLM 5.2 Thinking

via nano-gpt

What changed

Removed from the model catalog

repriced
Hy3

via opencode-go

What changed

  • nameHy3 (8x usage)Hy3
  • Input price$0.018/M$0.14/M▲ 700.0%
  • Output price$0.072/M$0.58/M▲ 700.0%
  • Cache read$0.0044/M$0.035/M▲ 700.0%
removed
DMind-1-Mini

via nano-gpt

What changed

Removed from the model catalog

capability
GLM-5.3

via merge-gateway

What changed

  • open weightsNoYesEnabled
removed
GLM 4.6V

via nano-gpt

What changed

Removed from the model catalog

removed
GLM-4 Flash

via nano-gpt

What changed

Removed from the model catalog

new model
GLM-5.3

via ollama-cloud

What changed

First observed in the model catalog

removed
Gemma 4 26B A4B Uncensored

via nano-gpt

What changed

Removed from the model catalog

removed
gpt-5.3-chat

via requesty

What changed

Removed from the model catalog

new model
GLM 5 Original Thinking

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3

via vivgrid

What changed

  • open weightsNoYesEnabled
repriced
Gemma 4 31B MeroMero v2 Thinking

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
removed
Jamba Large

via nano-gpt

What changed

Removed from the model catalog

capability
GLM 5.3 Flash Uncensored

via nano-gpt

What changed

  • attachmentNoYesEnabled
  • release date2026-08-272026-07-29
  • modalities.inputtexttext,image
new model
GLM 4.6V Original

via nano-gpt

What changed

First observed in the model catalog

new model
Abliterated Model

via nano-gpt

What changed

First observed in the model catalog

capability
Ornith 1.5 35B Thinking

via nano-gpt

What changed

  • release date2026-08-202026-07-29
capability
GLM-5.3

via opencode-go

What changed

  • open weightsNoYesEnabled
removed
Inflection 3 Productivity

via nano-gpt

What changed

Removed from the model catalog

removed
Opus Distill

via nano-gpt

What changed

Removed from the model catalog

repriced
Gemsicle

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
removed
Gemsicle

via nano-gpt

What changed

Removed from the model catalog

new model
Gemma 4 26B A4B Uncensored Thinking

via nano-gpt

What changed

First observed in the model catalog

new model
Dark Soul

via nano-gpt

What changed

First observed in the model catalog

capability
GLM-5.3

via ofox

What changed

  • open weightsNoYesEnabled
context
Inkling

via kilo

What changed

  • Output limit262K472K1.8×
new model
GLM 5.2

via nano-gpt

What changed

First observed in the model catalog

capability
Ornith 1.5 9B Thinking

via nano-gpt

What changed

  • release date2026-08-242026-07-29
repriced
Novelist

via nano-gpt

What changed

  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.33/M$0.45/M▲ 36.4%
  • Cache read$0.04/M$0.05/M▲ 25.0%
removed
GLM 4.7 Flash

via nano-gpt

What changed

Removed from the model catalog

repriced
Devstral 2

via kilo

What changed

  • Input price$0.44/M$0.4/M▼ 9.1%
  • Output price$2.20/M$2/M▼ 9.1%
  • Cache read$0.044/M$0.04/M▼ 9.1%
removed
GLM-4

via nano-gpt

What changed

Removed from the model catalog

removed
Auto

via trustedrouter

What changed

Removed from the model catalog

deprecated
Hy3 Free

via opencode

What changed

  • status—deprecated
repriced
Musica

via nano-gpt

What changed

  • Input price$0.08/M$0.12/M▲ 50.0%
  • Output price$0.33/M$0.38/M▲ 15.2%
  • Cache read$0.04/M$0.06/M▲ 50.0%
removed
GLM-5.3-Flash (EU)

via requesty

What changed

Removed from the model catalog

capability
Qwen 3.8 27B Obliterated

via nano-gpt

What changed

  • release date2026-08-242026-07-29
capability
Gemma 4 26B A4B MeroMero

via nano-gpt

What changed

  • release date2026-08-262026-07-29
removed
Novelist

via nano-gpt

What changed

Removed from the model catalog

new model
GLM 4.5 Air

via nano-gpt

What changed

First observed in the model catalog

removed
GLM 5.1 Thinking

via nano-gpt

What changed

Removed from the model catalog

new model
GLM-5.3

via friendli

What changed

First observed in the model catalog

37 events
repriced
DeepSeek V4 Flash (Fireworks AI)

via llmgateway-providers

What changed

  • Input price$0.14/M$0.22/M▲ 57.1%
  • Output price$0.28/M$0.66/M▲ 135.7%
  • Cache read$0.028/M$0.0070/M▼ 75.0%
context
DeepSeek V3.2

via kilo

What changed

  • Output limit66K147K2.3×
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.087/M$0.087/M▼ 0.3%
  • Output price$0.174/M$0.173/M▼ 0.3%
  • Cache read$0.017/M$0.017/M▼ 0.3%
repriced
GLM Latest

via openrouter

What changed

  • Output limit131K944K7.2×
  • Input price$1.25/M$1.20/M▼ 4.0%
  • Output price$4.40/M$4/M▼ 9.1%
  • Cache read$0.26/M$0.24/M▼ 7.7%
repriced
Z.ai: GLM Latest

via kilo

What changed

  • Context262K1.05M4×
  • Output limit131K944K7.2×
  • Input price$1.25/M$1.20/M▼ 4.0%
  • Output price$4.40/M$4/M▼ 9.1%
  • +1 more changes
removed
Kimi K2.6 (Tundra)

via llmgateway-providers

What changed

Removed from the model catalog

removed
Yi Lightning

via nano-gpt

What changed

Removed from the model catalog

context
DeepSeek V4 Flash 0731

via edenai

What changed

  • Context1.05M1.02M−2%
removed
AllenAI: Olmo 3 32B Think

via kilo

What changed

Removed from the model catalog

new model
DeepSeek V4 Pro 0813

via edenai

What changed

First observed in the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.083/M$0.14/M▲ 69.7%
  • Output price$0.33/M$0.58/M▲ 75.8%
  • Cache read$0.021/M$0.035/M▲ 69.7%
removed
Arcee AI: Virtuoso Large

via kilo

What changed

Removed from the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.742/M$0.731/M▼ 1.4%
  • Output price$1.48/M$1.46/M▼ 1.4%
  • Cache read$0.062/M$0.061/M▼ 1.4%
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
removed
Kimi K3 (Permafrost)

via llmgateway-providers

What changed

Removed from the model catalog

new model
Gemsicle

via nano-gpt

What changed

First observed in the model catalog

repriced
GPT OSS 20B

via edenai

What changed

  • Input price$0.03/M$0.02/M▼ 33.3%
  • Output price$0.13/M$0.1/M▼ 23.1%
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output price$0.1/M$0.14/M▲ 40.0%
  • Cache read$0.0070/M$0.01/M▲ 42.9%
new model
GLM-5.3

via crof

What changed

First observed in the model catalog

repriced
Qwen3.8 27B

via vercel

What changed

  • Input price$0.55/M$0.5/M▼ 9.1%
  • Output price$3.30/M$3/M▼ 9.1%
  • Cache read$0.11/M$0.1/M▼ 9.1%
  • cache write—$0.625/M
removed
Olmo 3 32B Think

via openrouter

What changed

Removed from the model catalog

capability
DeepSeek V4 Flash 0731

via edenai

What changed

  • tool callNoYesEnabled
repriced
DeepSeek V3.2

via openrouter

What changed

  • Output limit66K147K2.3×
  • Input price$0.269/M$0.26/M▼ 3.3%
  • Output price$0.4/M$0.38/M▼ 5.0%
  • Cache read$0.135/M$0.13/M▼ 3.3%
new model
GLM-5.3

via togetherai

What changed

First observed in the model catalog

new model
GLM-5.3 (SCX.ai)

via llmgateway-providers

What changed

First observed in the model catalog

capability
GPT OSS 120B

via edenai

What changed

  • tool callNoYesEnabled
capability
Llama-3.3-70B-Instruct

via edenai

What changed

  • tool callNoYesEnabled
new model
GLM5.3

via digitalocean

What changed

First observed in the model catalog

context
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit131K262K2×
repriced
GLM-5.3

via llmgateway

What changed

  • Input price$1.40/M$1.30/M▼ 7.1%
  • Output price$4.40/M$4/M▼ 9.1%
  • Cache read$0.26/M$0.25/M▼ 3.8%
context
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit131K262K2×
repriced
Gemma 4 31B IT

via kilo

What changed

  • Input price$0.08/M$0.07/M▼ 12.5%
  • Cache read$0.01/M$0.1/M▲ 900.0%
repriced
GLM-5.3

via deepinfra

What changed

  • Input price$1.40/M$1.20/M▼ 14.3%
  • Output price$4.40/M$4/M▼ 9.1%
  • Cache read$0.26/M$0.24/M▼ 7.7%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.05/M$0.045/M▼ 10.0%
  • Output price$0.1/M$0.09/M▼ 10.0%
  • Cache read$0.01/M$0.0090/M▼ 10.0%
repriced
Inkling

via openrouter

What changed

  • Output limit262K472K1.8×
  • Input price$0.95/M$1/M▲ 5.3%
  • Cache read$0.16/M$0.17/M▲ 6.3%
removed
Virtuoso Large

via openrouter

What changed

Removed from the model catalog

497 events
removed
Gemini 3.7 Flash (Iceberg)

via llmgateway-providers

What changed

Removed from the model catalog

new provider
DeepSeek V4 Flash 0731

via openreason

What changed

openreason began listing this model

new model
Kimi K2.7 Code

via neuralwatt

What changed

First observed in the model catalog

new provider
Kimi K2.7 Code

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

new model
Qwen3.8 Flash

via alibaba

What changed

First observed in the model catalog

capability
Seed 2.1 Turbo

via ofox

What changed

  • structured output—YesEnabled
context
Qwen3.8 2.4T A95B

via kilo

What changed

  • Context1.05M1M−5%
  • Output limit131K262K2×
new model
Dark Soul

via nano-gpt

What changed

First observed in the model catalog

capability
Qwen 3.8 27B Fable

via nano-gpt

What changed

  • tool callNoYesEnabled
new model
Grok 4.6

via kenari

What changed

First observed in the model catalog

new provider
MiniMax-M3

via vancine

What changed

vancine began listing this model

repriced
Claude Opus 4.1 (latest)

via requesty

What changed

  • Input price$13.50/M$15/M▲ 11.1%
  • Output price$67.50/M$75/M▲ 11.1%
  • Cache read$1.35/M$1.50/M▲ 11.1%
  • cache write$16.88/M$18.75/M▲ 11.1%
new model
Mistral Medium 3.5 (Free)

via kenari

What changed

First observed in the model catalog

repriced
Deepseek V4 Flash

via digitalocean

What changed

  • Input price$0.068/M$0.14/M▲ 106.2%
  • Output price$0.168/M$0.28/M▲ 66.7%
  • Cache read$0.017/M$0.028/M▲ 66.7%
new model
Muse Spark 1.2

via orcarouter

What changed

First observed in the model catalog

repriced
seed-1.8

via requesty

What changed

  • Input price$0.225/M$0.25/M▲ 11.1%
  • Output price$1.80/M$2/M▲ 11.1%
  • Cache read$0.045/M$0.05/M▲ 11.1%
new model
OrcaRouter Free

via orcarouter

What changed

First observed in the model catalog

repriced
Claude Opus 4.7

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$22.50/M$25/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$5.63/M$6.25/M▲ 11.1%
removed
Claude Sonnet 4 (latest)

via orcarouter

What changed

Removed from the model catalog

repriced
Claude Fable 5 (EU)

via requesty

What changed

  • Input price$9.90/M$11/M▲ 11.1%
  • Output price$49.50/M$55/M▲ 11.1%
  • Cache read$0.99/M$1.10/M▲ 11.1%
  • cache write$12.38/M$13.75/M▲ 11.1%
context
Seed 2.0 Mini

via volcengine

What changed

  • Output limit32K131K4.1×
new provider
GLM-5.3

via vancine

What changed

vancine began listing this model

context
GLM-5.3

via openrouter

What changed

  • Context1.05M1.31M1.3×
new model
Hy3

via orcarouter

What changed

First observed in the model catalog

new model
Gemini 3.6 Flash

via kenari

What changed

First observed in the model catalog

repriced
Nano Banana Pro

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$10.80/M$12/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
  • cache write$4.05/M$4.50/M▲ 11.1%
  • +1 more changes
repriced
devstral-latest

via requesty

What changed

  • Input price$0.396/M$0.44/M▲ 11.1%
  • Output price$1.98/M$2.20/M▲ 11.1%
  • Cache read$0.396/M$0.44/M▲ 11.1%
repriced
Kimi K2.6

via inceptron

What changed

  • Input price$0.54/M$0.53/M▼ 1.9%
  • Cache read$0.15/M$0.17/M▲ 13.3%
new provider
Kimi K3

via vancine

What changed

vancine began listing this model

new model
Hy3 (free)

via orcarouter

What changed

First observed in the model catalog

repriced
Claude Sonnet 5 (EU)

via requesty

What changed

  • Input price$1.98/M$2.20/M▲ 11.1%
  • Output price$9.90/M$11/M▲ 11.1%
  • Cache read$0.198/M$0.22/M▲ 11.1%
  • cache write$2.48/M$2.75/M▲ 11.1%
repriced
GLM Latest

via openrouter

What changed

  • Context1.05M1.31M1.3×
  • Output limit944K131K−86%
  • Input price$1.40/M$1.25/M▼ 10.7%
new model
GPT-5.6 Sol

via orcarouter

What changed

First observed in the model catalog

new model
GLM 5.3-Flash

via crof

What changed

First observed in the model catalog

repriced
GPT-4.1 (EU)

via requesty

What changed

  • Input price$1.98/M$2.20/M▲ 11.1%
  • Output price$7.92/M$8.80/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
repriced
Kimi K2.5

via orcarouter

What changed

  • Output limit262K33K−88%
  • cache write—$0/M
repriced
GLM 5.2 Short Flex

via neuralwatt

What changed

  • structured output—NoRemoved
  • Output limit200K32K−84%
  • Input price$0.725/M$0.943/M▲ 30.0%
  • Output price$2.25/M$2.92/M▲ 30.0%
  • +1 more changes
new model
DarkIdol

via nano-gpt

What changed

First observed in the model catalog

new model
Step 3.7 Flash

via kenari

What changed

First observed in the model catalog

repriced
Mistral Medium (latest) (EU)

via requesty

What changed

  • Input price$0.396/M$0.44/M▲ 11.1%
  • Output price$1.98/M$2.20/M▲ 11.1%
  • Cache read$0.396/M$0.44/M▲ 11.1%
repriced
DeepSeek Chat

via orcarouter

What changed

  • Input price$0.14/M$0.147/M▲ 5.0%
  • Output price$0.28/M$0.295/M▲ 5.4%
  • Cache read$0.028/M$0.02/M▼ 28.6%
new model
Wan v3.0 Video Prime

via vercel

What changed

First observed in the model catalog

new model
Gemini 3.5 Flash

via kenari

What changed

First observed in the model catalog

repriced
MiMo V2.5 Pro

via digitalocean

What changed

  • Input price$0.4/M$0.8/M▲ 100.0%
  • Output price$1.50/M$3/M▲ 100.0%
  • Cache read$0.08/M$0.16/M▲ 100.0%
removed
Kimi K2.6 Turbo

via fireworks-ai

What changed

Removed from the model catalog

repriced
Gemini 3.7 Flash

via requesty

What changed

  • Input price$0.6/M$0.75/M▲ 25.0%
  • Output price$3/M$3.75/M▲ 25.0%
  • Cache read$0.06/M$0.075/M▲ 25.0%
new model
Qwen3.8 Flash

via opencode-go

What changed

First observed in the model catalog

new model
Ling 3.0 Flash Fin (Free)

via vercel

What changed

First observed in the model catalog

new model
Gemini 3.1 Flash Lite

via orcarouter

What changed

First observed in the model catalog

new model
Hy4 preview

via openrouter

What changed

First observed in the model catalog

new provider
GPT OSS 120B

via openreason

What changed

openreason began listing this model

repriced
GPT 5.6 Terra Pro

via nano-gpt

What changed

  • Input price$1/M$2/M▲ 100.0%
  • Output price$6/M$12/M▲ 100.0%
  • Cache read$0.1/M$0.2/M▲ 100.0%
  • cache write$1.25/M$2.50/M▲ 100.0%
new provider
Kimi K2.6

via tokengo

What changed

tokengo began listing this model

new model
GLM-5.3-Flash

via togetherai

What changed

First observed in the model catalog

new model
Abliterated Model Large

via nano-gpt

What changed

First observed in the model catalog

new model
Kimi K2.7 Code (SCX.ai)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Qwen3.6 Plus

via requesty

What changed

  • Input price$0.45/M$0.5/M▲ 11.1%
  • Output price$2.70/M$3/M▲ 11.1%
  • Cache read$0.045/M$0.05/M▲ 11.1%
  • cache write$0.563/M$0.625/M▲ 11.1%
capability
Seed 2.1 Pro

via volcengine

What changed

  • structured output—YesEnabled
repriced
GPT-5.4 (EU)

via requesty

What changed

  • Input price$2.25/M$2.50/M▲ 11.1%
  • Output price$13.50/M$15/M▲ 11.1%
  • Cache read$0.225/M$0.25/M▲ 11.1%
  • tiers[object Object][object Object]
removed
Claude Opus 4 (latest)

via orcarouter

What changed

Removed from the model catalog

new model
Qwen3.8 Flash Next

via cortecs

What changed

First observed in the model catalog

repriced
Gemma 4 31B IT

via kilo

What changed

  • Input price$0.09/M$0.08/M▼ 11.1%
  • Output price$0.34/M$0.35/M▲ 2.9%
  • Cache read$0.05/M$0.01/M▼ 80.0%
context
GPT-5.5 Pro

via orcarouter

What changed

  • Output limit128K100K−22%
new model
Isometry

via nano-gpt

What changed

First observed in the model catalog

repriced
Seed 2.0 Pro

via requesty

What changed

  • Input price$0.45/M$0.5/M▲ 11.1%
  • Output price$2.70/M$3/M▲ 11.1%
  • Cache read$0.09/M$0.1/M▲ 11.1%
new model
OrcaRouter Fusion

via orcarouter

What changed

First observed in the model catalog

repriced
GPT-4.1 nano

via openrouter

What changed

  • Output limit33K943K28.8×
  • Cache read$0.025/M$0.03/M▲ 20.0%
repriced
Kimi K2.7 Code

via requesty

What changed

  • Input price$0.855/M$0.95/M▲ 11.1%
  • Output price$3.60/M$4/M▲ 11.1%
  • Cache read$0.171/M$0.19/M▲ 11.1%
removed
Kimi K3

via wandb

What changed

Removed from the model catalog

repriced
mistral-medium-3-5

via requesty

What changed

  • Input price$1.49/M$1.65/M▲ 11.1%
  • Output price$7.42/M$8.25/M▲ 11.1%
  • Cache read$1.49/M$1.65/M▲ 11.1%
repriced
GLM 5.3

via vercel

What changed

  • Output limit13K1M78.1×
  • Cache read$0.26/M$0.14/M▼ 46.2%
context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
repriced
Hy3

via kilo

What changed

  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • Cache read$0.033/M$0.035/M▲ 6.1%
context
Voxtral Small 24B 2507

via openrouter

What changed

  • Context32K33K1×
  • Output limit26K26K1×
new provider
Seed Evolving

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

removed
GPT-5.3 Chat (latest)

via orcarouter

What changed

Removed from the model catalog

removed
Qwen3.5 122B A10B TEE

via nano-gpt

What changed

Removed from the model catalog

new model
Hy4 preview

via tencent-token-plan

What changed

First observed in the model catalog

repriced
Gemini 3.1 Pro Preview Custom Tools

via orcarouter

What changed

  • Input price$4/M$2/M▼ 50.0%
  • Output price$18/M$12/M▼ 33.3%
repriced
Kimi K3

via llmgateway

What changed

  • Input price$3/M$2.83/M▼ 5.7%
  • Output price$15/M$14.13/M▼ 5.8%
  • Cache read$0.3/M$0.28/M▼ 6.7%
context
Qwen3.6 27B

via kilo

What changed

  • Output limit236K82K−65%
new model
GLM-5.3-Flash

via ofox

What changed

First observed in the model catalog

repriced
Claude Haiku 4.5 (latest)

via requesty

What changed

  • Input price$0.9/M$1/M▲ 11.1%
  • Output price$4.50/M$5/M▲ 11.1%
  • Cache read$0.09/M$0.1/M▲ 11.1%
  • cache write$1.13/M$1.25/M▲ 11.1%
capability
Seed 1.6 Flash

via volcengine

What changed

  • reasoningNoYesEnabled
new model
Gemini 3.1 Flash TTS Preview

via kenari

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via requesty

What changed

  • Input price$1.19/M$1.32/M▲ 11.1%
  • Output price$3.56/M$3.96/M▲ 11.1%
  • Cache read$0.04/M$0.044/M▲ 11.1%
repriced
ling-2.6-1t

via requesty

What changed

  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$2.25/M$2.50/M▲ 11.1%
capability
Seed 1.6 Flash

via ofox

What changed

  • structured output—YesEnabled
new model
Opus Distill

via nano-gpt

What changed

First observed in the model catalog

context
Kimi K2.6

via orcarouter

What changed

  • Output limit262K33K−88%
context
Seed 2.0 Lite

via volcengine

What changed

  • Output limit32K131K4.1×
repriced
Inkling

via requesty

What changed

  • Input price$1.68/M$1.87/M▲ 11.1%
  • Output price$4.21/M$4.68/M▲ 11.1%
  • Cache read$0.337/M$0.374/M▲ 11.1%
repriced
GLM 5.2 Short

via neuralwatt

What changed

  • structured output—NoRemoved
  • Output limit200K32K−84%
  • Cache read$0.362/M$0.145/M▼ 60.0%
repriced
mistral-medium-3-5@eu

via requesty

What changed

  • Input price$1.49/M$1.65/M▲ 11.1%
  • Output price$7.42/M$8.25/M▲ 11.1%
  • Cache read$1.49/M$1.65/M▲ 11.1%
new model
DeepSeek V4 Flash 0731

via orcarouter

What changed

First observed in the model catalog

removed
Gemini 3 Flash (Preview) (Iceberg)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
GPT-5.4 mini

via requesty

What changed

  • Input price$0.675/M$0.75/M▲ 11.1%
  • Output price$4.05/M$4.50/M▲ 11.1%
  • Cache read$0.068/M$0.075/M▲ 11.1%
removed
Qwen3.5 397B A17B FP8

via neuralwatt

What changed

Removed from the model catalog

repriced
Claude Opus 4.5 (latest)

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$22.50/M$25/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$5.63/M$6.25/M▲ 11.1%
repriced
DeepSeek V3.2

via openrouter

What changed

  • Output limit147K66K−56%
  • Input price$0.26/M$0.269/M▲ 3.5%
  • Output price$0.38/M$0.4/M▲ 5.3%
  • Cache read$0.13/M$0.135/M▲ 3.5%
repriced
GPT-4o mini (EU)

via requesty

What changed

  • Input price$0.148/M$0.165/M▲ 11.1%
  • Output price$0.594/M$0.66/M▲ 11.1%
  • Cache read$0.074/M$0.083/M▲ 11.1%
repriced
Gemma 4 31B IT

via orcarouter

What changed

  • Cache read—$0.02/M
new model
Grok 4.5

via orcarouter

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
GPT-5 Mini (EU)

via requesty

What changed

  • Input price$0.247/M$0.275/M▲ 11.1%
  • Output price$1.98/M$2.20/M▲ 11.1%
  • Cache read$0.025/M$0.028/M▲ 11.1%
new model
Qwen3.8 27B (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Claude Sonnet 4.5 (latest) (EU)

via requesty

What changed

  • Input price$2.97/M$3.30/M▲ 11.1%
  • Output price$14.85/M$16.50/M▲ 11.1%
  • Cache read$0.27/M$0.3/M▲ 11.1%
  • cache write$3.71/M$4.13/M▲ 11.1%
  • +1 more changes
repriced
Claw High

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
repriced
GPT-5.4 Pro

via requesty

What changed

  • Input price$27/M$30/M▲ 11.1%
  • Output price$162/M$180/M▲ 11.1%
  • Cache read$27/M$30/M▲ 11.1%
repriced
GPT-5.6 Terra

via github-copilot

What changed

  • cache write—$2.50/M
  • tiers[object Object][object Object]
context
Gemma 3 27B

via openrouter

What changed

  • Context262K131K−50%
repriced
kat-coder-pro

via requesty

What changed

  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$1.08/M$1.20/M▲ 11.1%
new model
GLM-5.3-Flash

via runinfra

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via requesty

What changed

  • Input price$0.126/M$0.14/M▲ 11.1%
  • Output price$0.252/M$0.28/M▲ 11.1%
  • Cache read$0.063/M$0.07/M▲ 11.1%
new model
GLM-5.3-Flash

via vivgrid

What changed

First observed in the model catalog

new model
Musica

via nano-gpt

What changed

First observed in the model catalog

new model
Qwen3.8 Flash (Alibaba Cloud)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Claude Haiku 4.5 (latest) (EU)

via requesty

What changed

  • Input price$0.99/M$1.10/M▲ 11.1%
  • Output price$4.95/M$5.50/M▲ 11.1%
  • Cache read$0.099/M$0.11/M▲ 11.1%
  • cache write$1.24/M$1.38/M▲ 11.1%
repriced
Qwen3.8 Max

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$5.40/M$6/M▲ 11.1%
  • Cache read$0.225/M$0.25/M▲ 11.1%
  • cache write$2.25/M$2.50/M▲ 11.1%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.05/M$0.06/M▲ 20.0%
  • Output price$0.1/M$0.12/M▲ 20.0%
  • Cache read$0.01/M$0.012/M▲ 20.0%
new model
Qwen3.8-Max

via modal

What changed

First observed in the model catalog

repriced
Google Gemini Flash Latest

via openrouter

What changed

  • Input price$0.375/M$0.75/M▲ 100.0%
  • Output price$1.88/M$3.75/M▲ 100.0%
  • Cache read$0.037/M$0.075/M▲ 100.0%
  • cache write$0.021/M$0.042/M▲ 100.0%
  • +1 more changes
capability
GLM-5.3

via kilo

What changed

  • structured outputNoYesEnabled
repriced
Claude Sonnet 4 (latest) (EU)

via requesty

What changed

  • Input price$2.70/M$3/M▲ 11.1%
  • Output price$13.50/M$15/M▲ 11.1%
  • Cache read$0.27/M$0.3/M▲ 11.1%
  • cache write$3.38/M$3.75/M▲ 11.1%
  • +1 more changes
new model
Voxtral Small (latest)

via edenai

What changed

First observed in the model catalog

new model
Qwen3.8 Flash

via alibaba-token-plan-cn

What changed

First observed in the model catalog

repriced
DeepSeek Reasoner

via orcarouter

What changed

  • Input price$0.435/M$0.147/M▼ 66.2%
  • Output price$0.87/M$0.295/M▼ 66.1%
removed
Kimi K2.6 Fast

via fireworks-ai

What changed

Removed from the model catalog

repriced
GLM-5.1 (EU)

via requesty

What changed

  • Input price$1.26/M$1.40/M▲ 11.1%
  • Output price$3.96/M$4.40/M▲ 11.1%
  • Cache read$1.26/M$1.40/M▲ 11.1%
new model
Gemini Robotics-ER 1.6 Preview

via orcarouter

What changed

First observed in the model catalog

new model
Kimi K2.7 Code

via orcarouter

What changed

First observed in the model catalog

repriced
GPT 5.6 Terra

via nano-gpt

What changed

  • Input price$1/M$2/M▲ 100.0%
  • Output price$6/M$12/M▲ 100.0%
  • Cache read$0.1/M$0.2/M▲ 100.0%
  • cache write$1.25/M$2.50/M▲ 100.0%
new model
Hy4 preview

via opencode-go

What changed

First observed in the model catalog

new provider
MiniMax-M3

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

new provider
DeepSeek V4 Flash

via vancine

What changed

vancine began listing this model

repriced
GPT-5 (EU)

via requesty

What changed

  • Input price$1.24/M$1.38/M▲ 11.1%
  • Output price$9.90/M$11/M▲ 11.1%
  • Cache read$0.124/M$0.138/M▲ 11.1%
repriced
Gemini 3.1 Flash Lite

via requesty

What changed

  • Input price$0.225/M$0.25/M▲ 11.1%
  • Output price$1.35/M$1.50/M▲ 11.1%
  • Cache read$0.022/M$0.025/M▲ 11.1%
  • cache write$0.075/M$0.083/M▲ 11.1%
  • +1 more changes
new model
MiniMax-M3

via orcarouter

What changed

First observed in the model catalog

new model
Gemini 3.1 Pro Preview

via kenari

What changed

First observed in the model catalog

context
MiniMax M3

via vercel

What changed

  • Context1M512K−49%
  • Output limit1M512K−49%
repriced
OpenAI GPT-oss-120b

via digitalocean

What changed

  • Input price$0.055/M$0.1/M▲ 81.8%
  • Output price$0.385/M$0.7/M▲ 81.8%
repriced
Gemini 3.6 Flash

via requesty

What changed

  • Input price$1.35/M$1.50/M▲ 11.1%
  • Output price$6.30/M$7/M▲ 11.1%
  • Cache read$0.135/M$0.15/M▲ 11.1%
new model
GLM-5.2

via orcarouter

What changed

First observed in the model catalog

repriced
glm-5.2-fast

via requesty

What changed

  • Input price$1.89/M$2.10/M▲ 11.1%
  • Output price$5.94/M$6.60/M▲ 11.1%
  • Cache read$0.189/M$0.21/M▲ 11.1%
repriced
Claw Low

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
new model
Step 3.7 Flash (Free)

via kenari

What changed

First observed in the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.0%
  • Output price$0.699/M$0.699/M▼ 0.0%
new provider
DeepSeek V4 Pro

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

new model
Qwen3.8 27B

via llmgateway

What changed

First observed in the model catalog

repriced
GPT-5.6 Luna (EU)

via requesty

What changed

  • Input price$0.198/M$0.22/M▲ 11.1%
  • Output price$1.19/M$1.32/M▲ 11.1%
  • Cache read$0.02/M$0.022/M▲ 11.1%
new model
DeepSeek V4 Pro 0813

via orcarouter

What changed

First observed in the model catalog

repriced
GPT-5.6 Luna

via github-copilot

What changed

  • cache write—$0.25/M
  • tiers[object Object][object Object]
repriced
GLM-5.3-Flash

via kilo

What changed

  • Input price$0.075/M$0.15/M▲ 100.0%
  • Output price$0.25/M$0.5/M▲ 100.0%
  • Cache read$0.015/M$0.03/M▲ 100.0%
new provider
Qwen3.8 Flash

via vancine

What changed

vancine began listing this model

removed
Kimi K2.6 Flex

via neuralwatt

What changed

Removed from the model catalog

repriced
Nemotron 3.5 Lightning 30B A3B

via openrouter

What changed

  • Output limit131K236K1.8×
  • Input price$0.08/M$0.1/M▲ 25.0%
  • Output price$0.2/M$0.25/M▲ 25.0%
  • Cache read$0.04/M$0.05/M▲ 25.0%
new model
Claude Opus 5

via kenari

What changed

First observed in the model catalog

capability
GLM-5.3

via openrouter

What changed

  • structured outputNoYesEnabled
capability
Seed 1.6 Vision

via ofox

What changed

  • structured output—YesEnabled
repriced
Nemotron 3.5 Lightning 30B A3B

via runinfra

What changed

  • Cache read—$0.01/M
repriced
GPT 5.6 Luna

via nano-gpt

What changed

  • Input price$0.1/M$0.2/M▲ 100.0%
  • Output price$0.6/M$1.20/M▲ 100.0%
  • Cache read$0.01/M$0.02/M▲ 100.0%
  • cache write$0.125/M$0.25/M▲ 100.0%
repriced
Hermes Low

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
removed
Gemini 3.6 Flash (Iceberg)

via llmgateway-providers

What changed

Removed from the model catalog

repriced
Hy3

via vercel

What changed

  • Context256K262K1×
  • Output limit128K262K2×
  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • +1 more changes
repriced
Qwen3.6 35B Fast

via neuralwatt

What changed

  • structured output—YesEnabled
  • Cache read$0.072/M$0.029/M▼ 60.0%
repriced
Grok 4.6

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$5.40/M$6/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$1.80/M$2/M▲ 11.1%
  • +1 more changes
repriced
MiMo-V2.5-Pro

via requesty

What changed

  • Input price$0.392/M$0.435/M▲ 11.1%
  • Output price$0.783/M$0.87/M▲ 11.1%
  • Cache read$0.0032/M$0.0036/M▲ 11.1%
removed
DeepSeek V4 Pro Lightning

via crof

What changed

Removed from the model catalog

context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
repriced
GLM 5.2 Fast

via neuralwatt

What changed

  • reasoningNoYesEnabled
  • structured output—NoRemoved
  • Cache read$0.362/M$0.145/M▼ 60.0%
new provider
MiniMax-M2.5

via tokengo

What changed

tokengo began listing this model

repriced
DeepSeek V4 Pro 0813

via requesty

What changed

  • Input price$1.19/M$1.32/M▲ 11.1%
  • Output price$3.56/M$3.96/M▲ 11.1%
  • Cache read$0.04/M$0.044/M▲ 11.1%
context
KAT-Coder-Pro V2.5

via openrouter

What changed

  • Context256K262K1×
new model
Kimi K3 Flex

via neuralwatt

What changed

First observed in the model catalog

repriced
Nano Banana 2

via requesty

What changed

  • Input price$0.45/M$0.5/M▲ 11.1%
  • Output price$1.80/M$2/M▲ 11.1%
  • tiers[object Object][object Object]
context
GPT-4.1 nano

via kilo

What changed

  • Output limit33K943K28.8×
repriced
Kimi K3 (EU)

via requesty

What changed

  • Input price$2.02/M$2.25/M▲ 11.1%
  • Output price$10.13/M$11.25/M▲ 11.1%
  • Cache read$0.203/M$0.225/M▲ 11.1%
new model
Abliterated Model

via nano-gpt

What changed

First observed in the model catalog

capability
Seed 1.6

via ofox

What changed

  • structured output—YesEnabled
repriced
ling-2.6-flash

via requesty

What changed

  • Input price$0.09/M$0.1/M▲ 11.1%
  • Output price$0.27/M$0.3/M▲ 11.1%
new model
GLM-5.3

via deepinfra

What changed

First observed in the model catalog

new model
Qwen3.8 Flash Next

via amd

What changed

First observed in the model catalog

repriced
GLM 5.2 Short Fast Flex

via neuralwatt

What changed

  • reasoningNoYesEnabled
  • structured output—NoRemoved
  • Output limit200K32K−84%
  • Input price$0.725/M$0.943/M▲ 30.0%
  • +2 more changes
repriced
Kimi K3

via openrouter

What changed

  • Input price$3/M$2.55/M▼ 15.0%
  • Output price$15/M$12.75/M▼ 15.0%
  • Cache read$0.3/M$0.256/M▼ 14.7%
new model
GLM-5.3-Flash (EU)

via requesty

What changed

First observed in the model catalog

new model
Qwen3.7 Flash

via orcarouter

What changed

First observed in the model catalog

repriced
MiniMax-M3

via requesty

What changed

  • Input price$0.24/M$0.3/M▲ 25.0%
  • Output price$0.96/M$1.20/M▲ 25.0%
  • Cache read$0.048/M$0.06/M▲ 25.0%
repriced
Gemini 3.5 Flash

via requesty

What changed

  • Input price$1.35/M$1.50/M▲ 11.1%
  • Output price$8.10/M$9/M▲ 11.1%
  • Cache read$0.135/M$0.15/M▲ 11.1%
  • cache write$1.42/M$1.58/M▲ 11.1%
repriced
Gemini 3.7 Flash (EU)

via requesty

What changed

  • Input price$0.66/M$0.825/M▲ 25.0%
  • Output price$3.30/M$4.13/M▲ 25.0%
  • Cache read$0.066/M$0.083/M▲ 25.0%
repriced
devstral-latest@eu

via requesty

What changed

  • Input price$0.396/M$0.44/M▲ 11.1%
  • Output price$1.98/M$2.20/M▲ 11.1%
  • Cache read$0.396/M$0.44/M▲ 11.1%
repriced
Claude Fable 5

via requesty

What changed

  • Input price$9/M$10/M▲ 11.1%
  • Output price$45/M$50/M▲ 11.1%
  • Cache read$0.9/M$1/M▲ 11.1%
  • cache write$11.25/M$12.50/M▲ 11.1%
capability
Seed Character

via volcengine

What changed

  • structured output—YesEnabled
new provider
GLM-5.3-Flash

via tokengo

What changed

tokengo began listing this model

removed
GPT-5-Codex

via orcarouter

What changed

Removed from the model catalog

repriced
Gemini 3.5 Flash Lite

via requesty

What changed

  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$2.25/M$2.50/M▲ 11.1%
  • Cache read$0.027/M$0.03/M▲ 11.1%
new model
Qwen3.7 Max

via orcarouter

What changed

First observed in the model catalog

new provider
Qwen3.8 Max

via vancine

What changed

vancine began listing this model

repriced
GLM 5.2 Short Fast

via neuralwatt

What changed

  • reasoningNoYesEnabled
  • structured output—NoRemoved
  • Output limit200K32K−84%
  • Cache read$0.362/M$0.145/M▼ 60.0%
removed
DeepSeek V4 Pro

via fireworks-ai

What changed

Removed from the model catalog

new model
GLM-5.3-Flash

via kenari

What changed

First observed in the model catalog

repriced
Kimi K2.6

via requesty

What changed

  • Input price$0.855/M$0.95/M▲ 11.1%
  • Output price$3.60/M$4/M▲ 11.1%
  • Cache read$0.144/M$0.16/M▲ 11.1%
repriced
Seed 2.0 Mini

via requesty

What changed

  • Input price$0.09/M$0.1/M▲ 11.1%
  • Output price$0.36/M$0.4/M▲ 11.1%
  • Cache read$0.018/M$0.02/M▲ 11.1%
repriced
GLM 5.2

via neuralwatt

What changed

  • structured output—NoRemoved
  • Cache read$0.362/M$0.145/M▼ 60.0%
context
GPT-5 Chat (latest)

via orcarouter

What changed

  • Output limit128K100K−22%
capability
Seed 2.0 Lite

via volcengine

What changed

  • structured output—YesEnabled
new model
Whisper Large v3 Turbo

via kenari

What changed

First observed in the model catalog

repriced
MiniMax-M2.7-highspeed

via requesty

What changed

  • Input price$0.54/M$0.6/M▲ 11.1%
  • Output price$2.16/M$2.40/M▲ 11.1%
  • Cache read$0.054/M$0.06/M▲ 11.1%
  • cache write$1.08/M$1.20/M▲ 11.1%
new model
Qwen3.8 27B

via regolo-ai

What changed

First observed in the model catalog

new model
Qwen3.8 Flash

via llmgateway

What changed

First observed in the model catalog

repriced
Claw Medium

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
new model
Hy3

via edenai

What changed

First observed in the model catalog

capability
GLM-5.3-Flash

via huggingface

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
repriced
MiniMax-M2.7

via requesty

What changed

  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$1.08/M$1.20/M▲ 11.1%
  • Cache read$0.054/M$0.06/M▲ 11.1%
  • cache write$1.08/M$1.20/M▲ 11.1%
repriced
DeepSeek V4 Pro

via orcarouter

What changed

  • Input price$0.56/M$0.442/M▼ 21.1%
  • Output price$1.12/M$0.884/M▼ 21.1%
  • Cache read$0.0036/M$0.06/M▲ 1555.2%
repriced
Claude Opus 4.8 (EU)

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$24.75/M$27.50/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
  • cache write$6.19/M$6.88/M▲ 11.1%
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Output limit944K384K−59%
  • Input price$1.12/M$1.32/M▲ 17.6%
  • Output price$3.37/M$3.96/M▲ 17.6%
  • Cache read$0.037/M$0.044/M▲ 17.6%
new model
Gemini 3.5 Flash

via orcarouter

What changed

First observed in the model catalog

new model
Novelist

via nano-gpt

What changed

First observed in the model catalog

new provider
Seed 2.0 Lite

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

new model
Qwen3 VL 235B A22B Instruct

via orcarouter

What changed

First observed in the model catalog

capability
Seed 2.0 Code

via edenai

What changed

  • tool callNoYesEnabled
  • structured outputNoYesEnabled
repriced
MiMo-V2.5

via requesty

What changed

  • Input price$0.126/M$0.14/M▲ 11.1%
  • Output price$0.252/M$0.28/M▲ 11.1%
  • Cache read$0.0025/M$0.0028/M▲ 11.1%
repriced
GLM-5

via orcarouter

What changed

  • Cache read$0.2/M$0.26/M▲ 30.0%
repriced
Kimi K2.7 Code

via llmgateway

What changed

  • Input price$0.95/M$0.89/M▼ 6.3%
  • Output price$4/M$3.71/M▼ 7.3%
  • Cache read$0.19/M$0.18/M▼ 5.3%
repriced
Step 3.7 Flash

via requesty

What changed

  • Input price$0.18/M$0.2/M▲ 11.1%
  • Output price$1.03/M$1.15/M▲ 11.1%
  • Cache read$0.036/M$0.04/M▲ 11.1%
repriced
DeepSeek V4 Flash (New)

via crof

What changed

  • Input price$0.12/M$0.08/M▼ 33.3%
  • Output price$0.21/M$0.1/M▼ 52.4%
removed
Kimi K2.6

via neuralwatt

What changed

Removed from the model catalog

context
GLM Latest

via openrouter

What changed

  • structured outputNoYesEnabled
  • Output limit131K944K7.2×
new model
Kimi K2.7 Code Fast

via neuralwatt

What changed

First observed in the model catalog

repriced
Gemini 3.7 Flash

via openrouter

What changed

  • Input price$0.375/M$0.75/M▲ 100.0%
  • Output price$1.88/M$3.75/M▲ 100.0%
  • Cache read$0.037/M$0.075/M▲ 100.0%
  • cache write$0.021/M$0.042/M▲ 100.0%
  • +1 more changes
removed
GPT OSS 20B

via edenai

What changed

Removed from the model catalog

new model
Tencent Hy4 Preview

via nano-gpt

What changed

First observed in the model catalog

capability
Seed 2.1 Turbo

via volcengine

What changed

  • structured output—YesEnabled
repriced
GPT-5.4

via orcarouter

What changed

  • Input price$5/M$2.50/M▼ 50.0%
  • Output price$22.50/M$15/M▼ 33.3%
repriced
GLM-5.2

via digitalocean

What changed

  • Input price$0.7/M$1.40/M▲ 100.0%
  • Output price$2.20/M$4.40/M▲ 100.0%
  • Cache read$0.105/M$0.21/M▲ 100.0%
repriced
Kimi K2.7 Code

via vercel

What changed

  • Cache read$0.19/M$0.16/M▼ 15.8%
repriced
GPT-5.1 (EU)

via requesty

What changed

  • Input price$1.24/M$1.38/M▲ 11.1%
  • Output price$9.90/M$11/M▲ 11.1%
  • Cache read$0.124/M$0.138/M▲ 11.1%
repriced
GPT-5.6 Luna

via requesty

What changed

  • Input price$0.18/M$0.2/M▲ 11.1%
  • Output price$1.08/M$1.20/M▲ 11.1%
  • Cache read$0.018/M$0.02/M▲ 11.1%
  • tiers[object Object][object Object]
new provider
Seed 2.1 Turbo

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.466/M$0.466/M▼ 0.0%
  • Output price$0.932/M$0.931/M▼ 0.0%
new model
GLM-5.3 Flash (SCX.ai)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Kimi K2.7 Code Flex

via neuralwatt

What changed

  • Context262K262K−0%
  • Output limit262K262K−0%
  • Input price$0.475/M$0.618/M▲ 30.0%
  • Output price$2/M$2.60/M▲ 30.0%
  • +1 more changes
repriced
Claude Opus 4.6

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$22.50/M$25/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$5.63/M$6.25/M▲ 11.1%
new model
Qwen3.6 35B

via neuralwatt

What changed

First observed in the model catalog

new provider
DeepSeek V3.2

via tokengo

What changed

tokengo began listing this model

new model
Claude Sonnet 5

via orcarouter

What changed

First observed in the model catalog

repriced
Gemini 3.5 Flash Lite (EU)

via requesty

What changed

  • Input price$0.297/M$0.33/M▲ 11.1%
  • Output price$2.48/M$2.75/M▲ 11.1%
  • Cache read$0.03/M$0.033/M▲ 11.1%
repriced
GPT-5.6 Sol (EU)

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$29.70/M$33/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
repriced
Kimi K3

via requesty

What changed

  • Input price$2.02/M$2.25/M▲ 11.1%
  • Output price$10.13/M$11.25/M▲ 11.1%
  • Cache read$0.203/M$0.225/M▲ 11.1%
removed
Inflection 3 Pi

via nano-gpt

What changed

Removed from the model catalog

context
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit131K262K2×
new model
Glm 5.3

via cloudflare-workers-ai

What changed

First observed in the model catalog

repriced
nvidia-nemotron-3-super-120b-a12b

via requesty

What changed

  • Input price$0.09/M$0.1/M▲ 11.1%
  • Output price$0.45/M$0.5/M▲ 11.1%
repriced
thinkingcap-qwen3.6-27b

via requesty

What changed

  • Input price$0.36/M$0.4/M▲ 11.1%
  • Output price$2.70/M$3/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
removed
Nemotron 3 Ultra 550B A55B

via edenai

What changed

Removed from the model catalog

new model
Gembrain

via nano-gpt

What changed

First observed in the model catalog

repriced
GLM 5.2 Flex

via neuralwatt

What changed

  • structured output—NoRemoved
  • Input price$0.725/M$0.943/M▲ 30.0%
  • Output price$2.25/M$2.92/M▲ 30.0%
  • Cache read$0.181/M$0.094/M▼ 48.0%
repriced
Grok Build 0.1

via requesty

What changed

  • Input price$0.9/M$1/M▲ 11.1%
  • Output price$1.80/M$2/M▲ 11.1%
  • Cache read$0.09/M$0.1/M▲ 11.1%
new model
Qwen3.8 Flash

via crossmodel

What changed

First observed in the model catalog

new model
Qwen 3.8 27B Fable

via nano-gpt

What changed

First observed in the model catalog

new model
GLM-5.3

via cortecs

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.12/M$0.11/M▼ 8.3%
  • Output price$0.42/M$0.408/M▼ 2.9%
  • cache write$0.06/M$0.055/M▼ 8.3%
repriced
Claude Opus 4.5 (latest) (EU)

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$24.75/M$27.50/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
  • cache write$6.19/M$6.88/M▲ 11.1%
new model
Qwen3.6 Flash

via orcarouter

What changed

First observed in the model catalog

new model
DeepSeek V4 Pro 0813

via scnet-token-plan

What changed

First observed in the model catalog

repriced
Mistral Small 4 (EU)

via requesty

What changed

  • Input price$0.148/M$0.165/M▲ 11.1%
  • Output price$0.594/M$0.66/M▲ 11.1%
  • Cache read$0.148/M$0.165/M▲ 11.1%
new model
Tencent Hy4 Preview

via vercel

What changed

First observed in the model catalog

new provider
DeepSeek V4 Pro

via vancine

What changed

vancine began listing this model

repriced
Hy3

via requesty

What changed

  • Input price$0.126/M$0.14/M▲ 11.1%
  • Output price$0.522/M$0.58/M▲ 11.1%
  • Cache read$0.032/M$0.035/M▲ 11.1%
new model
Grok Imagine Image 2.0

via kenari

What changed

First observed in the model catalog

repriced
Kimi K2.6 (EU)

via requesty

What changed

  • Input price$0.855/M$0.95/M▲ 11.1%
  • Output price$3.60/M$4/M▲ 11.1%
  • Cache read$0.855/M$0.95/M▲ 11.1%
repriced
Qwen3.6 27B

via openrouter

What changed

  • Output limit236K82K−65%
  • Input price$0.6/M$0.32/M▼ 46.7%
  • Output price$3.60/M$3.20/M▼ 11.1%
  • Cache read$0.12/M—
new model
Shadow Siren

via nano-gpt

What changed

First observed in the model catalog

context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
context
Claude Sonnet 4.5 (latest)

via orcarouter

What changed

  • Context200K1M5×
new model
Garnet

via nano-gpt

What changed

First observed in the model catalog

repriced
GPT-5.6 Terra

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$10.80/M$12/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
  • tiers[object Object][object Object]
new provider
DeepSeek V4 Flash

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

removed
Qwen3.5 397B Fast

via neuralwatt

What changed

Removed from the model catalog

new model
GPT OSS 120B

via orcarouter

What changed

First observed in the model catalog

new model
Qwen3.8 27B

via orcarouter

What changed

First observed in the model catalog

repriced
Claude Opus 5

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$22.50/M$25/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$5.63/M$6.25/M▲ 11.1%
repriced
Mistral Small 4

via requesty

What changed

  • Input price$0.148/M$0.165/M▲ 11.1%
  • Output price$0.594/M$0.66/M▲ 11.1%
  • Cache read$0.148/M$0.165/M▲ 11.1%
new model
Gemini 3.5 Flash Lite

via orcarouter

What changed

First observed in the model catalog

context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Context1.05M1.05M1×
  • Output limit944K384K−59%
repriced
MiniMax-M3 (EU)

via requesty

What changed

  • Input price$0.36/M$0.4/M▲ 11.1%
  • Output price$1.80/M$2/M▲ 11.1%
  • Cache read$0.09/M$0.1/M▲ 11.1%
repriced
Meta: Llama 4 Maverick

via kilo

What changed

  • Output price$0.696/M$0.8/M▲ 14.9%
new provider
Kimi K2.7 Code

via openreason

What changed

openreason began listing this model

repriced
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit262K131K−50%
  • Cache read$0.25/M$0.2/M▼ 20.0%
new model
Qwen3.8 Max

via orcarouter

What changed

First observed in the model catalog

new model
Hy3 (Free)

via kenari

What changed

First observed in the model catalog

context
Mistral: Voxtral Small 24B 2507

via kilo

What changed

  • Context32K33K1×
  • Output limit26K26K1×
new provider
GLM-5.1

via tokengo

What changed

tokengo began listing this model

removed
Kimi K2.6 Fast

via neuralwatt

What changed

Removed from the model catalog

repriced
gpt-5.3-chat

via requesty

What changed

  • Input price$1.57/M$1.75/M▲ 11.1%
  • Output price$12.60/M$14/M▲ 11.1%
  • Cache read$0.158/M$0.175/M▲ 11.1%
repriced
Hy3

via openrouter

What changed

  • Input price$0.083/M$0.132/M▲ 60.0%
  • Output price$0.33/M$0.528/M▲ 60.0%
  • Cache read$0.021/M$0.033/M▲ 60.0%
new provider
GLM-5

via tokengo

What changed

tokengo began listing this model

repriced
Qwen3.8 27B

via openrouter

What changed

  • Input price$0.425/M$0.4/M▼ 5.9%
  • Cache read$0.085/M$0.05/M▼ 41.2%
  • cache write$0.531/M—
capability
Seed 2.0 Mini

via ofox

What changed

  • structured output—YesEnabled
repriced
nemotron-3-nano-omni@eu

via requesty

What changed

  • Input price$0.054/M$0.06/M▲ 11.1%
  • Output price$0.216/M$0.24/M▲ 11.1%
  • Cache read$0.054/M$0.06/M▲ 11.1%
new provider
GLM-5.3

via volcengine-coding-plan

What changed

volcengine-coding-plan began listing this model

repriced
Muse Glimmer 30B

via openrouter

What changed

  • Output limit118K16K−86%
  • Input price$0.35/M$0.3/M▼ 14.3%
  • Output price$1.50/M$1.20/M▼ 20.0%
repriced
Qwen3.8 2.4T A95B

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$5.40/M$6/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
new model
Fabled

via nano-gpt

What changed

First observed in the model catalog

removed
Kimi K2.7 Code Fast

via fireworks-ai

What changed

Removed from the model catalog

repriced
Qwen3.8 27B

via crof

What changed

  • Input price$0.25/M$0.2/M▼ 20.0%
  • Output price$2.10/M$1.50/M▼ 28.6%
  • Cache read$0.06/M$0.03/M▼ 50.0%
repriced
GLM-5.1

via requesty

What changed

  • Input price$1.26/M$1.40/M▲ 11.1%
  • Output price$3.96/M$4.40/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
removed
DeepSeek V4 Flash

via fireworks-ai

What changed

Removed from the model catalog

repriced
Deepseek V4 Pro

via digitalocean

What changed

  • Input price$0.87/M$1.74/M▲ 100.0%
  • Output price$1.74/M$3.48/M▲ 100.0%
  • Cache read$0.174/M$0.348/M▲ 100.0%
repriced
GPT-5.5 (EU)

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$27/M$30/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • tiers[object Object][object Object]
new model
GLM-5.3 (EU)

via requesty

What changed

First observed in the model catalog

new model
GLM-5.3 (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.408/M$0.404/M▼ 1.0%
  • Output price$1.51/M$1.50/M▼ 1.1%
  • cache write$0.204/M$0.202/M▼ 1.0%
repriced
Kimi K3

via digitalocean

What changed

  • Input price$2.85/M$3/M▲ 5.3%
  • Output price$14.25/M$15/M▲ 5.3%
  • Cache read$0.285/M$0.3/M▲ 5.3%
new model
Tencent: Hy4 preview

via kilo

What changed

First observed in the model catalog

removed
Gemini 3.1 Pro (Preview) (Iceberg)

via llmgateway-providers

What changed

Removed from the model catalog

capability
Hy4 preview

via openrouter

What changed

  • family—Hy
  • open weightsNoYesEnabled
new provider
DeepSeek V4 Flash

via tokengo

What changed

tokengo began listing this model

new model
Qwen3.8 Flash

via edenai

What changed

First observed in the model catalog

repriced
Fugu Ultra

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$27/M$30/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • tiers[object Object][object Object]
new model
Claude Opus 4.8

via orcarouter

What changed

First observed in the model catalog

repriced
Claude Sonnet 4.6 (EU)

via requesty

What changed

  • Input price$2.97/M$3.30/M▲ 11.1%
  • Output price$14.85/M$16.50/M▲ 11.1%
  • Cache read$0.27/M$0.3/M▲ 11.1%
  • cache write$3.71/M$4.13/M▲ 11.1%
removed
Kimi K2.5

via neuralwatt

What changed

Removed from the model catalog

repriced
GLM-5.3

via requesty

What changed

  • Input price$1.26/M$1.40/M▲ 11.1%
  • Output price$3.96/M$4.40/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
repriced
Gemini 3.6 Flash

via github-copilot

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
new model
Qwen3.8 Flash

via alibaba-token-plan

What changed

First observed in the model catalog

repriced
Grok 4.5

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$5.40/M$6/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$1.80/M$2/M▲ 11.1%
  • +1 more changes
new model
GLM-5.3

via huggingface

What changed

First observed in the model catalog

repriced
OpenAI GPT-5.6 Sol

via digitalocean

What changed

  • Input price$4/M$5/M▲ 25.0%
  • Output price$20/M$30/M▲ 50.0%
  • Cache read$0.4/M$0.5/M▲ 25.0%
  • tiers[object Object][object Object]
context
Nemotron 3.5 Lightning 30B A3B

via kilo

What changed

  • Output limit131K236K1.8×
capability
Seed Evolving

via volcengine

What changed

  • structured output—YesEnabled
removed
GPT-5.1 Codex Max

via orcarouter

What changed

Removed from the model catalog

capability
Seed 1.8

via volcengine

What changed

  • structured output—YesEnabled
new model
Qwen3.8 Flash

via alibaba-cn

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via vercel

What changed

  • Context1.05M1M−5%
  • Output limit1.05M384K−63%
  • Input price$1.74/M$0.66/M▼ 62.1%
  • Output price$3.48/M$1.98/M▼ 43.1%
  • +1 more changes
repriced
DeepSeek V4 Flash 0731 (EU)

via requesty

What changed

  • Input price$0.126/M$0.14/M▲ 11.1%
  • Output price$0.252/M$0.28/M▲ 11.1%
  • Cache read$0.063/M$0.07/M▲ 11.1%
repriced
grok-4.2-beta

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$5.40/M$6/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
  • cache write$1.80/M$2/M▲ 11.1%
  • +1 more changes
repriced
Deepseek 3.2

via digitalocean

What changed

  • Input price$0.25/M$0.5/M▲ 100.0%
  • Output price$0.8/M$1.60/M▲ 100.0%
  • Cache read$0.075/M$0.15/M▲ 100.0%
repriced
MiniMax-M2.5-highspeed

via orcarouter

What changed

  • Cache read$0.06/M$0.03/M▼ 50.0%
context
Qwen: Qwen3 VL 30B A3B Instruct

via kilo

What changed

  • Context131K262K2×
  • Output limit33K16K−50%
removed
Qwen3.6 35B A3B

via neuralwatt

What changed

Removed from the model catalog

new model
DeepSeek V4 Pro 0813

via nvidia

What changed

First observed in the model catalog

repriced
thinkingcap-qwen3.6-27b@eu

via requesty

What changed

  • Input price$0.36/M$0.4/M▲ 11.1%
  • Output price$2.70/M$3/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
repriced
GLM 5.3 Preview Thinking

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
repriced
DeepSeek V4 Pro (EU)

via requesty

What changed

  • Input price$1.57/M$1.75/M▲ 11.1%
  • Output price$3.15/M$3.50/M▲ 11.1%
  • Cache read$0.396/M$0.44/M▲ 11.1%
repriced
Mistral Medium (latest)

via requesty

What changed

  • Input price$0.396/M$0.44/M▲ 11.1%
  • Output price$1.98/M$2.20/M▲ 11.1%
  • Cache read$0.396/M$0.44/M▲ 11.1%
repriced
Kimi K2.7 Code (EU)

via requesty

What changed

  • Input price$1.13/M$1.25/M▲ 11.1%
  • Output price$4.05/M$4.50/M▲ 11.1%
  • Cache read$0.279/M$0.31/M▲ 11.1%
capability
Hy4 preview

via kilo

What changed

  • nameTencent: Hy4 previewHy4 preview
  • family—Hy
  • open weightsNoYesEnabled
new provider
DeepSeek V4 Pro

via tokengo

What changed

tokengo began listing this model

new model
GLM5.3 Flash

via digitalocean

What changed

First observed in the model catalog

repriced
Gemini 2.5 Flash (EU)

via requesty

What changed

  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$2.25/M$2.50/M▲ 11.1%
  • Cache read$0.068/M$0.075/M▲ 11.1%
  • cache write$0.495/M$0.55/M▲ 11.1%
repriced
GPT-5.6 Sol

via requesty

What changed

  • Input price$3.60/M$4/M▲ 11.1%
  • Output price$18/M$20/M▲ 11.1%
  • Cache read$0.36/M$0.4/M▲ 11.1%
  • cache write$4.50/M$5/M▲ 11.1%
  • +1 more changes
repriced
ring-2.6-1t

via requesty

What changed

  • Input price$0.27/M$0.3/M▲ 11.1%
  • Output price$2.25/M$2.50/M▲ 11.1%
new model
GLM-5.3-Flash

via cortecs

What changed

First observed in the model catalog

new model
Hy4 preview

via tencent-tokenhub

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
Grok 4.3

via requesty

What changed

  • Input price$1.13/M$1.25/M▲ 11.1%
  • Output price$2.25/M$2.50/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
  • cache write$1.13/M$1.25/M▲ 11.1%
  • +1 more changes
repriced
Hermes Medium

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
repriced
Gemma 4 31B

via neuralwatt

What changed

  • Cache read$0.036/M$0.014/M▼ 60.0%
new model
Chimera X

via nano-gpt

What changed

First observed in the model catalog

new provider
GLM-5.3-Flash

via vancine

What changed

vancine began listing this model

context
DeepSeek V3.2

via kilo

What changed

  • Output limit147K66K−56%
repriced
Gemini 2.5 Pro

via orcarouter

What changed

  • Input price$2.50/M$1.25/M▼ 50.0%
  • Output price$15/M$10/M▼ 33.3%
repriced
Gemma 4 26B A4B IT

via requesty

What changed

  • Input price$0.063/M$0.07/M▲ 11.1%
  • Output price$0.306/M$0.34/M▲ 11.1%
  • Cache read$0.063/M$0.07/M▲ 11.1%
new provider
Qwen3.5 397B-A17B

via tokengo

What changed

tokengo began listing this model

new provider
DeepSeek-V3.1

via tokengo

What changed

tokengo began listing this model

new model
GLM 5.3 Flash

via wandb

What changed

First observed in the model catalog

new model
Mistral Large (Free)

via kenari

What changed

First observed in the model catalog

new model
Ornith 1.5 35B A3B

via runinfra

What changed

First observed in the model catalog

repriced
Qwen3 VL 30B A3B Instruct

via openrouter

What changed

  • Output limit33K16K−50%
  • Input price$0.13/M$0.15/M▲ 15.4%
  • Output price$0.52/M$0.6/M▲ 15.4%
new model
Ling 3.0 Flash Fin

via vercel

What changed

First observed in the model catalog

new model
Hy3

via kenari

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via orcarouter

What changed

  • Cache read—$0.0075/M
repriced
GLM 5.2

via inceptron

What changed

  • Input price$0.75/M$0.71/M▼ 5.3%
  • Output price$2.40/M$2.35/M▼ 2.1%
  • Cache read$0.17/M$0.12/M▼ 29.4%
new model
DeepSeek V4 Pro 0813

via wandb

What changed

First observed in the model catalog

capability
Seed 2.0 Pro

via volcengine

What changed

  • structured output—YesEnabled
repriced
Hermes High

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
capability
Seed Evolving

via ofox

What changed

  • structured output—YesEnabled
new model
MiniMax-M2.7-highspeed

via kenari

What changed

First observed in the model catalog

new model
GLM 5.3

via fireworks-ai

What changed

First observed in the model catalog

capability
Seed 2.1 Pro

via ofox

What changed

  • structured output—YesEnabled
repriced
Llama 4 Maverick

via digitalocean

What changed

  • Input price$0.2/M$0.25/M▲ 25.0%
  • Output price$0.696/M$0.87/M▲ 25.0%
repriced
Gemini 3.1 Pro Preview

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$10.80/M$12/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
  • cache write$4.05/M$4.50/M▲ 11.1%
  • +1 more changes
repriced
Claude Opus 4.8

via requesty

What changed

  • Input price$4.50/M$5/M▲ 11.1%
  • Output price$22.50/M$25/M▲ 11.1%
  • Cache read$0.45/M$0.5/M▲ 11.1%
  • cache write$5.63/M$6.25/M▲ 11.1%
removed
Claude Opus 4.1 (latest)

via orcarouter

What changed

Removed from the model catalog

repriced
Qwen3.7 Plus

via requesty

What changed

  • Input price$0.28/M$0.32/M▲ 14.3%
  • Output price$1.12/M$1.28/M▲ 14.3%
  • Cache read$0.028/M$0.032/M▲ 14.3%
  • cache write$0.35/M$0.4/M▲ 14.3%
repriced
GPT 5.6 Luna Pro

via nano-gpt

What changed

  • Input price$0.1/M$0.2/M▲ 100.0%
  • Output price$0.6/M$1.20/M▲ 100.0%
  • Cache read$0.01/M$0.02/M▲ 100.0%
  • cache write$0.125/M$0.25/M▲ 100.0%
removed
Gemini 3 Pro Preview

via orcarouter

What changed

Removed from the model catalog

new model
GLM-5.3-Flash

via ollama-cloud

What changed

First observed in the model catalog

new model
GLM 5.3

via baseten

What changed

First observed in the model catalog

removed
GPT OSS 20B

via fireworks-ai

What changed

Removed from the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.0%
  • Output price$0.757/M$0.757/M▼ 0.0%
removed
MiniMax-M2.7

via fireworks-ai

What changed

Removed from the model catalog

new model
Kimi K3 (SCX.ai)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
GPT-5.4

via requesty

What changed

  • Input price$2.48/M$2.75/M▲ 11.1%
  • Output price$14.85/M$16.50/M▲ 11.1%
  • Cache read$0.247/M$0.275/M▲ 11.1%
repriced
nemotron-3-ultra-nvfp4

via requesty

What changed

  • Input price$0.54/M$0.6/M▲ 11.1%
  • Output price$2.16/M$2.40/M▲ 11.1%
  • Cache read$0.108/M$0.12/M▲ 11.1%
capability
Seed Character

via ofox

What changed

  • structured output—YesEnabled
new model
OrcaRouter Fusion Mini

via orcarouter

What changed

First observed in the model catalog

new model
MiniMax-M2.7

via kenari

What changed

First observed in the model catalog

repriced
Gemini 3.1 Flash Lite (EU)

via requesty

What changed

  • Input price$0.247/M$0.275/M▲ 11.1%
  • Output price$1.49/M$1.65/M▲ 11.1%
  • Cache read$0.025/M$0.028/M▲ 11.1%
  • cache write$0.082/M$0.092/M▲ 11.1%
  • +1 more changes
repriced
inkling-256k

via requesty

What changed

  • Input price$1.50/M$1.87/M▲ 25.0%
  • Output price$3.74/M$4.68/M▲ 25.0%
  • Cache read$0.299/M$0.374/M▲ 25.0%
new model
Grok 4.6

via orcarouter

What changed

First observed in the model catalog

removed
Qwen3.6 27B

via regolo-ai

What changed

Removed from the model catalog

repriced
GPT-5.6 Sol

via github-copilot

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
  • +1 more changes
repriced
GPT-5.5 Pro

via requesty

What changed

  • Input price$27/M$30/M▲ 11.1%
  • Output price$162/M$180/M▲ 11.1%
new model
Muse Spark 1.1

via orcarouter

What changed

First observed in the model catalog

new model
GLM 5.3 Flash

via modal

What changed

First observed in the model catalog

capability
Seed 1.6

via volcengine

What changed

  • structured output—YesEnabled
repriced
nvidia-nemotron-3-ultra

via requesty

What changed

  • Input price$0.45/M$0.5/M▲ 11.1%
  • Output price$2.25/M$2.50/M▲ 11.1%
context
Muse Glimmer 30B

via kilo

What changed

  • Output limit118K16K−86%
new provider
Kimi K3

via tokengo

What changed

tokengo began listing this model

repriced
Claude Sonnet 5

via requesty

What changed

  • Input price$1.80/M$2/M▲ 11.1%
  • Output price$9/M$10/M▲ 11.1%
  • Cache read$0.18/M$0.2/M▲ 11.1%
  • cache write$2.25/M$2.50/M▲ 11.1%
repriced
Claude Opus 4.7 (EU)

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$24.75/M$27.50/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
  • cache write$6.19/M$6.88/M▲ 11.1%
repriced
Qwen3.5 35B-A3B

via requesty

What changed

  • Input price$0.126/M$0.14/M▲ 11.1%
  • Output price$0.9/M$1/M▲ 11.1%
  • Cache read$0.045/M$0.05/M▲ 11.1%
repriced
Claude Opus 5 (EU)

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$24.75/M$27.50/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
  • cache write$6.19/M$6.88/M▲ 11.1%
context
Qwen3.5 122B-A10B

via kilo

What changed

  • Output limit236K82K−65%
repriced
qwen3.5-2b

via requesty

What changed

  • Input price$0.018/M$0.02/M▲ 11.1%
  • Output price$0.09/M$0.1/M▲ 11.1%
new model
Ling 3.0 Flash Fin Free

via opencode

What changed

First observed in the model catalog

repriced
Gemini Flash Latest

via orcarouter

What changed

  • Input price$1.50/M$0.5/M▼ 66.7%
  • Output price$9/M$3/M▼ 66.7%
  • Cache read$0.15/M$0.1/M▼ 33.3%
  • input audio$1.50/M—
new model
Kimi K3

via orcarouter

What changed

First observed in the model catalog

repriced
Qwen3.5 27B

via requesty

What changed

  • Input price$0.234/M$0.26/M▲ 11.1%
  • Output price$2.34/M$2.60/M▲ 11.1%
new model
GLM 5.3 Flash TEE

via nano-gpt

What changed

First observed in the model catalog

repriced
Claude Sonnet 4.5 (latest)

via requesty

What changed

  • Input price$2.70/M$3/M▲ 11.1%
  • Output price$13.50/M$15/M▲ 11.1%
  • Cache read$0.27/M$0.3/M▲ 11.1%
  • cache write$3.38/M$3.75/M▲ 11.1%
  • +1 more changes
repriced
Llama 4 Maverick 17B Instruct

via hyper

What changed

  • Input price$0.274/M$0.284/M▲ 3.6%
  • Output price$0.899/M$0.934/M▲ 3.9%
  • cache write$0.137/M$0.142/M▲ 3.6%
repriced
nemotron-lightning-3.5-30b-a3b

via requesty

What changed

  • Input price$0.045/M$0.05/M▲ 11.1%
  • Output price$0.18/M$0.2/M▲ 11.1%
  • Cache read$0.0090/M$0.01/M▲ 11.1%
new model
DeepSeek V4 Flash Vision Exp

via orcarouter

What changed

First observed in the model catalog

new model
GPT-5.6 Terra

via orcarouter

What changed

First observed in the model catalog

repriced
Claude Sonnet 4.6

via requesty

What changed

  • Input price$2.70/M$3/M▲ 11.1%
  • Output price$13.50/M$15/M▲ 11.1%
  • Cache read$0.27/M$0.3/M▲ 11.1%
  • cache write$3.38/M$3.75/M▲ 11.1%
repriced
Nemotron 3.5 Lightning 30B

via vercel

What changed

  • Output price$0.15/M$0.2/M▲ 33.3%
  • Cache read$0.05/M$0.01/M▼ 80.0%
repriced
DeepSeek V4 Flash

via orcarouter

What changed

  • Input price$0.19/M$0.147/M▼ 22.6%
  • Output price$0.37/M$0.295/M▼ 20.3%
  • Cache read$0.0028/M$0.02/M▲ 614.3%
repriced
GPT-5.3 Codex

via requesty

What changed

  • Input price$1.57/M$1.75/M▲ 11.1%
  • Output price$12.60/M$14/M▲ 11.1%
  • Cache read$0.158/M$0.175/M▲ 11.1%
new model
Qwen3.8 Max

via kenari

What changed

First observed in the model catalog

repriced
GLM-5.2

via requesty

What changed

  • Input price$1.08/M$1.20/M▲ 11.1%
  • Output price$3.78/M$4.20/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
repriced
Seed 2.0 Code

via requesty

What changed

  • Input price$0.45/M$0.5/M▲ 11.1%
  • Output price$2.70/M$3/M▲ 11.1%
  • Cache read$0.09/M$0.1/M▲ 11.1%
repriced
nemotron-3-nano-omni

via requesty

What changed

  • Input price$0.054/M$0.06/M▲ 11.1%
  • Output price$0.216/M$0.24/M▲ 11.1%
  • Cache read$0.054/M$0.06/M▲ 11.1%
new model
Gemini 3.7 Flash

via kenari

What changed

First observed in the model catalog

new model
Moonlight Dusk

via nano-gpt

What changed

First observed in the model catalog

new model
Gemini 3.6 Flash

via orcarouter

What changed

First observed in the model catalog

capability
Seed 2.0 Lite

via ofox

What changed

  • structured output—YesEnabled
repriced
GPT-5.5

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$29.70/M$33/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
new model
OrcaRouter Fusion Flash

via orcarouter

What changed

First observed in the model catalog

repriced
Claude Opus 4.6 (EU)

via requesty

What changed

  • Input price$4.95/M$5.50/M▲ 11.1%
  • Output price$24.75/M$27.50/M▲ 11.1%
  • Cache read$0.495/M$0.55/M▲ 11.1%
  • cache write$6.19/M$6.88/M▲ 11.1%
new model
DeepSeek V4 Flash (free)

via orcarouter

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via digitalocean

What changed

  • Input price$0.08/M$0.14/M▲ 75.0%
  • Output price$0.252/M$0.28/M▲ 11.1%
  • Cache read$0.025/M$0.028/M▲ 11.1%
new model
GLM-5.3

via orcarouter

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via crossmodel

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.075/M$0.089/M▲ 18.3%
  • Output price$0.15/M$0.177/M▲ 18.3%
  • Cache read$0.015/M$0.018/M▲ 18.3%
new model
Qwen3.7 Plus

via orcarouter

What changed

First observed in the model catalog

new model
Claude Fable 5

via orcarouter

What changed

First observed in the model catalog

repriced
Qwen3.5 122B-A10B

via openrouter

What changed

  • Output limit236K82K−65%
  • Input price$0.26/M$0.29/M▲ 11.5%
  • Output price$2.08/M$2.40/M▲ 15.4%
capability
Seed 1.8

via ofox

What changed

  • structured output—YesEnabled
repriced
GPT-5.4 Pro

via orcarouter

What changed

  • Input price$60/M$30/M▼ 50.0%
  • Output price$270/M$180/M▼ 33.3%
repriced
o4-mini (EU)

via requesty

What changed

  • Input price$1.09/M$1.21/M▲ 11.1%
  • Output price$4.36/M$4.84/M▲ 11.1%
  • Cache read$0.272/M$0.302/M▲ 11.1%
repriced
Qwen3.7 Max

via requesty

What changed

  • Input price$2.25/M$2.50/M▲ 11.1%
  • Output price$6.75/M$7.50/M▲ 11.1%
  • Cache read$0.225/M$0.25/M▲ 11.1%
  • cache write$2.81/M$3.13/M▲ 11.1%
new model
GLM-5.3

via kenari

What changed

First observed in the model catalog

repriced
Google Gemini Flash Latest

via kilo

What changed

  • Input price$0.375/M$0.75/M▲ 100.0%
  • Output price$1.88/M$3.75/M▲ 100.0%
  • Cache read$0.037/M$0.075/M▲ 100.0%
  • cache write$0.021/M$0.042/M▲ 100.0%
  • +1 more changes
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
Qwen3.8 2.4T A95B

via vercel

What changed

  • Output limit131K128K−2%
  • Cache read$0.2/M$0.25/M▲ 25.0%
repriced
GLM-5.3-Flash

via requesty

What changed

  • Input price$0.135/M$0.15/M▲ 11.1%
  • Output price$0.45/M$0.5/M▲ 11.1%
  • Cache read$0.027/M$0.03/M▲ 11.1%
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.29/M$1.31/M▲ 1.9%
  • Output price$4.22/M$4.27/M▲ 1.1%
  • cache write$0.645/M$0.657/M▲ 1.9%
repriced
DeepSeek V4 Flash

via requesty

What changed

  • Input price$0.396/M$0.44/M▲ 11.1%
  • Output price$1.19/M$1.32/M▲ 11.1%
  • Cache read$0.013/M$0.014/M▲ 11.1%
new model
Claude Sonnet 4.6

via kenari

What changed

First observed in the model catalog

repriced
GPT-5 Nano (EU)

via requesty

What changed

  • Input price$0.05/M$0.055/M▲ 11.1%
  • Output price$0.396/M$0.44/M▲ 11.1%
  • Cache read$0.0050/M$0.0055/M▲ 11.1%
repriced
GLM 5.3 Preview

via nano-gpt

What changed

  • Input price$1.40/M$1/M▼ 28.6%
  • Output price$4.40/M$3.20/M▼ 27.3%
  • Cache read$0.26/M$0.2/M▼ 23.1%
repriced
Gemini 3.1 Pro Preview

via orcarouter

What changed

  • Input price$4/M$2/M▼ 50.0%
  • Output price$18/M$12/M▼ 33.3%
  • input audio—$2/M
removed
Kimi K2.7 Code

via neuralwatt

What changed

Removed from the model catalog

capability
Seed 2.0 Mini

via volcengine

What changed

  • structured output—YesEnabled
repriced
GLM-5.3-Flash

via llmgateway

What changed

  • Input price$0.15/M$0.13/M▼ 13.3%
  • Output price$0.5/M$0.4/M▼ 20.0%
  • Cache read$0.03/M$0.024/M▼ 20.0%
repriced
GPT-4.1 mini (EU)

via requesty

What changed

  • Input price$0.396/M$0.44/M▲ 11.1%
  • Output price$1.58/M$1.76/M▲ 11.1%
  • Cache read$0.099/M$0.11/M▲ 11.1%
repriced
GLM-5.2 (EU)

via requesty

What changed

  • Input price$1.08/M$1.20/M▲ 11.1%
  • Output price$3.78/M$4.20/M▲ 11.1%
  • Cache read$0.234/M$0.26/M▲ 11.1%
repriced
GPT-4.1 nano (EU)

via requesty

What changed

  • Input price$0.099/M$0.11/M▲ 11.1%
  • Output price$0.396/M$0.44/M▲ 11.1%
  • Cache read$0.025/M$0.028/M▲ 11.1%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▼ 0.0%
  • Output price$1.05/M$1.05/M▼ 0.0%
removed
GLM 5.3 Flash

via wandb

What changed

Removed from the model catalog

removed
Lyria 3 Clip Preview

via edenai

What changed

Removed from the model catalog

new provider
GLM-5.2

via tokengo

What changed

tokengo began listing this model

capability
Seed 1.6 Vision

via volcengine

What changed

  • reasoningNoYesEnabled
new model
Claude Opus 5

via orcarouter

What changed

First observed in the model catalog

repriced
Z.ai: GLM Latest

via kilo

What changed

  • structured outputNoYesEnabled
  • Output limit131K944K7.2×
  • Input price$1.40/M$1.20/M▼ 14.3%
  • Output price$4.40/M$1.20/M▼ 72.7%
  • +1 more changes
new provider
GLM-5.3

via tokengo

What changed

tokengo began listing this model

context
Qwen3.8 27B

via kilo

What changed

  • Context1M262K−74%
removed
Kimi K2.5 Fast

via neuralwatt

What changed

Removed from the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.757/M$0.757/M▼ 0.0%
  • Output price$0.757/M$0.757/M▼ 0.0%
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.87/M$0.772/M▼ 11.3%
  • Output price$1.74/M$1.54/M▼ 11.3%
  • Cache read$0.072/M$0.064/M▼ 11.3%
repriced
GPT-5.4 nano

via requesty

What changed

  • Input price$0.18/M$0.2/M▲ 11.1%
  • Output price$1.13/M$1.25/M▲ 11.1%
  • Cache read$0.018/M$0.02/M▲ 11.1%
repriced
Gemini 3.5 Flash (EU)

via requesty

What changed

  • Input price$1.49/M$1.65/M▲ 11.1%
  • Output price$8.91/M$9.90/M▲ 11.1%
  • Cache read$0.148/M$0.165/M▲ 11.1%
  • cache write$1.57/M$1.74/M▲ 11.1%
removed
Gemma 3 27B TEE

via nano-gpt

What changed

Removed from the model catalog

repriced
GPT-5.6 Terra (EU)

via requesty

What changed

  • Input price$1.98/M$2.20/M▲ 11.1%
  • Output price$11.88/M$13.20/M▲ 11.1%
  • Cache read$0.198/M$0.22/M▲ 11.1%
new model
DeepSeek V4 Pro (0813)

via crof

What changed

First observed in the model catalog

new model
Luminous Mirror

via nano-gpt

What changed

First observed in the model catalog

capability
Seed 2.0 Mini

via edenai

What changed

  • tool callNoYesEnabled
  • structured outputNoYesEnabled
removed
GPT OSS 20B

via edenai

What changed

Removed from the model catalog

new model
Qwen3 VL 235B A22B Thinking

via orcarouter

What changed

First observed in the model catalog

new model
GPT-5.6 Luna

via orcarouter

What changed

First observed in the model catalog

new model
Qwen3.8 Flash (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

capability
Seed 2.0 Pro

via ofox

What changed

  • structured output—YesEnabled
new model
Qwen3.5 Flash

via orcarouter

What changed

First observed in the model catalog

new model
Qwen3.8 27B (free)

via orcarouter

What changed

First observed in the model catalog

288 events
repriced
GLM-5

via alibaba-cn

What changed

  • Input price$0.86/M$0.573/M▼ 33.4%
  • Output price$3.15/M$2.58/M▼ 18.1%
  • tiers[object Object]
capability
GPT 5.3 Codex (Fast)

via vercel

What changed

  • temperatureNoYesEnabled
new model
DeepSeek V4 Flash (Consensus Protocol)

via llmgateway-providers

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via requesty

What changed

First observed in the model catalog

capability
GPT-5.3 Codex

via opper

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via openai

What changed

  • temperatureNoYesEnabled
new model
DeepSeek V4 Pro

via neuralwatt

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via huggingface

What changed

First observed in the model catalog

capability
GPT-5.4 Mini

via merge-gateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via ofox

What changed

  • temperatureNoYesEnabled
repriced
Qwen3 235B A22B Instruct 2507

via openrouter

What changed

  • Output limit16K236K14.4×
  • Input price$0.09/M$0.087/M▼ 2.8%
  • Output price$0.55/M$0.35/M▼ 36.4%
  • Cache read—$0.018/M
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.188/M$0.178/M▼ 5.3%
  • Output price$0.7/M$0.68/M▼ 2.9%
  • cache write$0.094/M$0.089/M▼ 5.3%
capability
GPT-5.2

via impossibl

What changed

  • temperatureNoYesEnabled
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.758/M$0.757/M▼ 0.2%
  • Output price$0.758/M$0.757/M▼ 0.2%
capability
GPT-5.3 Codex

via requesty

What changed

  • temperatureNoYesEnabled
repriced
Qwen3.8 Flash

via openrouter

What changed

  • Input price$0.16/M$0.15/M▼ 6.3%
new model
DeepSeek V4 Flash 0731

via volcengine

What changed

First observed in the model catalog

capability
GPT-5.2

via databricks

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 Nano

via merge-gateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via azure-cognitive-services

What changed

  • temperatureNoYesEnabled
capability
GPT 5.4 (Fast)

via vercel

What changed

  • temperatureNoYesEnabled
new model
GLM-5.3-Flash

via llmgateway

What changed

First observed in the model catalog

new model
GLM-5.2

via volcengine

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via deepinfra

What changed

First observed in the model catalog

context
GLM-5.3-Flash

via openrouter

What changed

  • structured outputNoYesEnabled
  • Context1.05M1.31M1.3×
capability
GPT-5.4 mini

via orcarouter

What changed

  • temperatureNoYesEnabled
repriced
GLM-5.1

via alibaba-cn

What changed

  • Input price$0.87/M$0.825/M▼ 5.2%
  • Output price$3.48/M$3.30/M▼ 5.1%
  • tiers[object Object]
capability
GPT 5.1

via nano-gpt

What changed

  • temperatureNoYesEnabled
repriced
Qwen3.6 35B-A3B

via kilo

What changed

  • Input price$0.14/M$0.1/M▼ 28.6%
  • Output price$1/M$0.9/M▼ 10.0%
new model
GLM 5.3 Flash

via baseten

What changed

First observed in the model catalog

capability
GPT-5.4

via crossmodel

What changed

  • temperatureNoYesEnabled
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▼ 0.2%
  • Output price$1.05/M$1.05/M▼ 0.2%
capability
GPT-5.4 mini

via xpersona

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via llmgateway

What changed

  • temperatureNoYesEnabled
capability
GPT 5.4 Nano

via nano-gpt

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via github-copilot

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via edenai

What changed

  • temperatureNoYesEnabled
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output price$0.075/M$0.1/M▲ 33.3%
capability
GPT-5.4

via nearai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 Mini (Azure)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
repriced
Hermes Low

via nano-gpt

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,image,video,pdftext
  • Output limit66K131K2×
  • Input price$0.25/M$1.40/M▲ 460.0%
  • +2 more changes
new model
Qwen 3.8 Flash

via vercel

What changed

First observed in the model catalog

capability
GPT-5.4 nano

via fastrouter

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 (EU)

via requesty

What changed

  • temperatureNoYesEnabled
context
DeepSeek V4 Flash 0731

via hyper

What changed

  • Context1.05M1M−5%
capability
GPT-5.4

via merge-gateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via ofox

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via orcarouter

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via llmgateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via kenari

What changed

  • temperatureNoYesEnabled
capability
Grok 4.3

via vercel

What changed

  • structured output—YesEnabled
  • release date2026-04-302026-04-17
  • last updated2026-04-302026-04-17
capability
GPT-5.1

via cloudflare-ai-gateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via venice

What changed

  • temperatureNoYesEnabled
removed
Rocinante 12B

via openrouter

What changed

Removed from the model catalog

capability
GPT-5.4 nano

via requesty

What changed

  • temperatureNoYesEnabled
new model
GLM 5.3 Flash

via venice

What changed

First observed in the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • Cache read$0.033/M$0.035/M▲ 6.1%
capability
GPT-5.4 mini

via impossibl

What changed

  • temperatureNoYesEnabled
repriced
Kimi K2.6

via inceptron

What changed

  • Input price$0.56/M$0.54/M▼ 3.6%
  • Cache read$0.17/M$0.13/M▼ 23.5%
capability
GPT 5.4 Mini (Fast)

via vercel

What changed

  • temperatureNoYesEnabled
repriced
Qwen3.8 Flash

via kilo

What changed

  • Input price$0.16/M$0.15/M▼ 6.3%
repriced
MiniMax: MiniMax M2.7 (free)

via kilo

What changed

  • Output limit177K197K1.1×
  • Cache read—$0/M
  • reasoning—$0/M
capability
GPT-5.4

via model-oracle-ai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via snowflake-cortex

What changed

  • temperatureNoYesEnabled
new model
DeepSeek V4 Flash Vision Exp (DeepSeek)

via llmgateway-providers

What changed

First observed in the model catalog

capability
GPT-5.1

via pioneer

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 Nano (Azure)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via ai-router

What changed

  • temperatureNoYesEnabled
context
Nemotron 3 Ultra 550B A55B

via kilo

What changed

  • Context512K262K−49%
  • Output limit461K16K−96%
capability
GPT-5.4 Nano

via azure

What changed

  • temperatureNoYesEnabled
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
capability
Gemma 4 26B A4B Uncensored TEE

via nano-gpt

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
new model
Gemini 3.1 Flash Lite (EU)

via edenai

What changed

First observed in the model catalog

capability
GPT-5.1 (EU)

via requesty

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via edenai

What changed

  • temperatureNoYesEnabled
new model
Ling 3.0 Flash Fin (free)

via kilo

What changed

First observed in the model catalog

repriced
OpenAI GPT-5.6 Sol

via digitalocean

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • tiers[object Object][object Object]
capability
gpt-5.4

via sap-ai-core

What changed

  • temperatureNoYesEnabled
new model
GLM 5.3 Flash

via vercel

What changed

First observed in the model catalog

context
Llama 3.1 8B

via wandb

What changed

  • Context128K131K1×
  • Output limit128K131K1×
capability
GPT-5.3 Codex

via fastrouter

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via orcarouter

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via orcarouter

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via openai

What changed

  • temperatureNoYesEnabled
new model
GPT OSS 120B (EU)

via edenai

What changed

First observed in the model catalog

removed
Ministral 8B

via openrouter

What changed

Removed from the model catalog

removed
Hy4 preview

via tencent-token-plan

What changed

Removed from the model catalog

new model
Qwen3.8 Flash

via hyper

What changed

First observed in the model catalog

capability
GPT-5.4

via databricks

What changed

  • temperatureNoYesEnabled
removed
Hy4 preview

via tencent-tokenhub

What changed

Removed from the model catalog

capability
Grok Imagine Video 1.5

via vercel

What changed

  • temperatureYesNoRemoved
  • release date2026-06-222026-05-30
  • last updated2026-06-222026-05-30
new model
Gemini 3.5 Transcribe

via vercel

What changed

First observed in the model catalog

capability
GPT-5.2

via unorouter

What changed

  • temperatureNoYesEnabled
repriced
DeepSeek V4 Flash

via neuralwatt

What changed

  • Input price$0.104/M$0.14/M▲ 34.6%
  • Output price$0.207/M$0.28/M▲ 35.3%
  • Cache read$0.026/M$0.028/M▲ 7.7%
removed
Ox Alpha

via venice

What changed

Removed from the model catalog

capability
GPT-5.4 Mini

via azure-cognitive-services

What changed

  • temperatureNoYesEnabled
capability
gpt-5.2

via sap-ai-core

What changed

  • temperatureNoYesEnabled
capability
Seed 1.8

via ofox

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
repriced
Hy3

via openrouter

What changed

  • Input price$0.132/M$0.083/M▼ 37.5%
  • Output price$0.528/M$0.33/M▼ 37.5%
  • Cache read$0.033/M$0.021/M▼ 37.5%
capability
GPT-5.1

via abacus

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via unorouter

What changed

  • temperatureNoYesEnabled
capability
GPT 5.2 (Fast)

via vercel

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via venice

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via databricks

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via orcarouter

What changed

  • temperatureNoYesEnabled
repriced
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
capability
GPT-5.1

via snowflake-cortex

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex (Azure)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
repriced
Kimi K2.7 Code

via openrouter

What changed

  • Input price$0.67/M$0.66/M▼ 1.5%
  • Cache read$0.19/M$0.18/M▼ 5.3%
repriced
Kimi K2.5

via hyper

What changed

  • Input price$0.544/M$0.544/M▼ 0.1%
  • Output price$2.85/M$2.76/M▼ 3.3%
  • cache write$0.272/M$0.272/M▼ 0.1%
capability
GPT-5.4 Mini (OpenAI)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
removed
Laguna M.1

via nano-gpt

What changed

Removed from the model catalog

repriced
Nemotron 3 Ultra 550B A55B

via openrouter

What changed

  • Context512K262K−49%
  • Output limit461K16K−96%
  • Input price$0.6/M$0.5/M▼ 16.7%
  • Output price$3.60/M$2.20/M▼ 38.9%
  • +1 more changes
repriced
Qwen3-Coder 480B-A35B Instruct

via alibaba

What changed

  • tiers[object Object],[object Object]
capability
GPT-5.4 mini

via ofox

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via edenai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via cloudflare-ai-gateway

What changed

  • temperatureNoYesEnabled
new model
DeepSeek V4 Flash Flex

via neuralwatt

What changed

First observed in the model catalog

new model
Qwen3.8 Flash

via openrouter

What changed

First observed in the model catalog

context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.2%
  • Output price$0.7/M$0.699/M▼ 0.2%
capability
GPT-5.1 (OpenAI)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
GPT-5 Chat (latest)

via orcarouter

What changed

  • temperatureYesNoRemoved
repriced
Kimi K2.7 Code

via inceptron

What changed

  • Input price$0.67/M$0.66/M▼ 1.5%
  • Cache read$0.19/M$0.18/M▼ 5.3%
capability
GPT-5.2

via edenai

What changed

  • temperatureNoYesEnabled
repriced
Qwen3-Coder 30B-A3B Instruct

via alibaba-cn

What changed

  • tiers[object Object],[object Object]
capability
GPT-5.3 Codex (OpenAI)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via llmgateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via daoxe

What changed

  • temperatureNoYesEnabled
repriced
Qwen3.5 397B-A17B

via alibaba-cn

What changed

  • Input price$0.43/M$0.172/M▼ 60.0%
  • Output price$2.58/M$1.03/M▼ 60.0%
  • reasoning$2.58/M$1.03/M▼ 60.0%
  • tiers[object Object]
capability
GPT-5.4

via unorouter

What changed

  • temperatureNoYesEnabled
context
Qwen3.6 27B

via kilo

What changed

  • Output limit82K236K2.9×
capability
GPT-5.4 mini

via github-copilot

What changed

  • temperatureNoYesEnabled
new model
Ling 3.0 Flash Fin (free)

via openrouter

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via zhipuai

What changed

First observed in the model catalog

repriced
Qwen3.6 35B-A3B

via openrouter

What changed

  • Input price$0.14/M$0.1/M▼ 28.6%
  • Output price$1/M$0.9/M▼ 10.0%
capability
GPT-5.4 nano

via ofox

What changed

  • temperatureNoYesEnabled
new model
DeepSeek V4 Flash Vision Exp

via llmgateway

What changed

First observed in the model catalog

capability
GPT-5.4

via opper

What changed

  • temperatureNoYesEnabled
repriced
GLM-5.3 Flash

via merge-gateway

What changed

  • Input price$0.075/M$0.015/M▼ 80.0%
  • Output price$0.25/M$0.05/M▼ 80.0%
  • Cache read$0.015/M$0.0030/M▼ 80.0%
repriced
DeepSeek V4 Flash

via llmgateway

What changed

  • Input price$0.076/M$0.051/M▼ 32.9%
  • Output price$0.153/M$0.104/M▼ 32.0%
  • Cache read$0.014/M$0.0097/M▼ 30.7%
repriced
Hermes Medium

via nano-gpt

What changed

  • Context205K1.05M5.1×
  • Input limit205K1.05M5.1×
  • Input price$0.315/M$1.40/M▲ 344.4%
  • Output price$1.26/M$4.40/M▲ 249.2%
  • +1 more changes
capability
GPT-5.4

via edenai

What changed

  • temperatureNoYesEnabled
repriced
MiMo-V2.5-Pro

via kilo

What changed

  • Input price$1/M$0.435/M▼ 56.5%
  • Output price$3/M$0.87/M▼ 71.0%
  • Cache read$0.2/M$0.0040/M▼ 98.0%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.467/M$0.466/M▼ 0.2%
  • Output price$0.934/M$0.932/M▼ 0.2%
new model
GLM 5.3 Flash

via empiriolabs

What changed

First observed in the model catalog

repriced
Qwen3-Coder 30B-A3B Instruct

via alibaba

What changed

  • tiers[object Object],[object Object]
capability
GPT-5.4 mini

via edenai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via github-copilot

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via llmgateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via cloudflare-ai-gateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via cortecs

What changed

  • temperatureNoYesEnabled
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.081/M$0.08/M▼ 1.9%
  • Output price$0.162/M$0.159/M▼ 1.9%
  • Cache read$0.016/M$0.016/M▼ 1.9%
capability
Seed 1.8

via volcengine

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
new model
GLM 5.3 Flash Uncensored

via nano-gpt

What changed

First observed in the model catalog

repriced
Nemotron 3 Nano 30B A3B

via openrouter

What changed

  • Output limit236K228K−3%
  • Cache read$0.03/M$0.025/M▼ 16.7%
new model
Muse Image 1.0

via vercel

What changed

First observed in the model catalog

capability
GPT-5.1

via nearai

What changed

  • temperatureNoYesEnabled
repriced
GLM-4.6

via openrouter

What changed

  • Output limit131K16K−88%
  • Input price$0.5/M$0.43/M▼ 14.0%
  • Output price$2/M$1.75/M▼ 12.5%
  • Cache read$0.1/M$0.08/M▼ 20.0%
capability
GPT-5.4 mini

via opper

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via abacus

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via freemodel

What changed

  • temperatureNoYesEnabled
context
Qwen3.5 35B A3B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
repriced
Qwen3-Coder 480B-A35B Instruct

via alibaba-cn

What changed

  • tiers[object Object],[object Object]
repriced
GLM-5

via hyper

What changed

  • Input price$0.91/M$0.92/M▲ 1.1%
  • Output price$2.81/M$2.98/M▲ 5.8%
  • cache write$0.455/M$0.46/M▲ 1.1%
capability
GPT-5.4 nano

via openai

What changed

  • temperatureNoYesEnabled
repriced
GLM-4.6

via kilo

What changed

  • Context203K198K−2%
  • Output limit131K16K−88%
  • Input price$0.5/M$0.43/M▼ 14.0%
  • Output price$2/M$1.75/M▼ 12.5%
  • +1 more changes
capability
GPT-5.3 Codex

via impossibl

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via freemodel

What changed

  • temperatureNoYesEnabled
new model
Gemini 3.5 Transcribe Live

via vercel

What changed

First observed in the model catalog

context
Qwen: Qwen3 235B A22B Instruct 2507

via kilo

What changed

  • Output limit16K236K14.4×
removed
Baichuan M2 32B Medical

via nano-gpt

What changed

Removed from the model catalog

capability
GPT-5.1

via ofox

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via orcarouter

What changed

  • temperatureNoYesEnabled
new model
Hy4 preview

via tencent-token-plan

What changed

First observed in the model catalog

context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
capability
GLM-5.3-Flash

via kilo

What changed

  • structured outputNoYesEnabled
new model
Qwen3.8 27B

via hyper

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash (Gonka24)

via llmgateway-providers

What changed

  • Input price$0.075/M$0.051/M▼ 32.0%
  • Output price$0.175/M$0.104/M▼ 40.6%
  • Cache read$0.015/M$0.0097/M▼ 37.4%
new model
Qwen3.8 Flash

via nano-gpt

What changed

First observed in the model catalog

capability
GPT-5.4 Nano

via azure-cognitive-services

What changed

  • temperatureNoYesEnabled
new model
Gemini 3.7 Flash (Iceberg)

via llmgateway-providers

What changed

First observed in the model catalog

capability
GPT-5.3 Codex

via venice

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via pioneer

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via azure

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via impossibl

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via amazon-bedrock

What changed

  • temperatureNoYesEnabled
new model
Hy4 preview

via tencent-tokenhub

What changed

First observed in the model catalog

capability
GPT-5.4

via freemodel

What changed

  • temperatureNoYesEnabled
capability
Grok Imagine Image 2.0

via vercel

What changed

  • temperatureYesNoRemoved
capability
Seed 1.6 Flash

via ofox

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
capability
GPT-5.1 (Azure)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
Grok Build 0.1

via vercel

What changed

  • structured output—YesEnabled
  • release date2026-05-202026-04-16
  • last updated2026-05-202026-04-16
capability
GPT-5.4 nano

via nearai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via anyapi

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2

via abacus

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via merge-gateway

What changed

  • temperatureNoYesEnabled
context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
capability
GPT-5.4

via cortecs

What changed

  • temperatureNoYesEnabled
repriced
Qwen3.8 Flash

via hyper

What changed

  • Input price$0.16/M$0.15/M▼ 6.3%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.06/M$0.05/M▼ 16.7%
  • Output price$0.12/M$0.1/M▼ 16.7%
  • Cache read$0.012/M$0.01/M▼ 16.7%
repriced
Claw Low

via nano-gpt

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,image,video,pdftext
  • Output limit66K131K2×
  • Input price$0.25/M$1.40/M▲ 460.0%
  • +2 more changes
capability
GPT-5.2 (Azure)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 Nano (OpenAI)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
new model
GLM-5.3-Flash

via zai-coding-plan

What changed

First observed in the model catalog

repriced
GLM 5.3 Flash

via venice

What changed

  • Input price$0.094/M$0.15/M▲ 60.0%
  • Output price$0.313/M$0.5/M▲ 60.0%
  • Cache read$0.019/M$0.03/M▲ 60.0%
repriced
Qwen3.5 35B-A3B

via openrouter

What changed

  • Output limit236K66K−72%
  • Input price$0.25/M$0.225/M▼ 10.0%
  • Output price$1.25/M$1.80/M▲ 44.0%
  • Cache read$0.25/M$0.225/M▼ 10.0%
new model
GLM-5.3

via zhipuai

What changed

First observed in the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.2%
  • Output price$0.758/M$0.757/M▼ 0.2%
capability
GPT-5.3 Codex

via edenai

What changed

  • temperatureNoYesEnabled
new model
DeepSeek V4 Pro 0813

via volcengine

What changed

First observed in the model catalog

capability
GPT-5.4

via snowflake-cortex

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via fastrouter

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via pioneer

What changed

  • temperatureNoYesEnabled
repriced
Qwen3.6 27B

via openrouter

What changed

  • Output limit82K236K2.9×
  • Input price$0.32/M$0.6/M▲ 87.5%
  • Output price$3.20/M$3.60/M▲ 12.5%
  • Cache read—$0.12/M
capability
GPT-5.4

via opencode

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via pioneer

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 (Azure)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via openai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via ofox

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 mini

via cloudflare-ai-gateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via llmgateway

What changed

  • temperatureNoYesEnabled
repriced
MiniMax: MiniMax M3 (free)

via kilo

What changed

  • modalities.inputtext,image,videotext,image
  • Output limit944K1.05M1.1×
  • Cache read—$0/M
  • reasoning—$0/M
repriced
Claw High

via nano-gpt

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,image,pdftext
  • Context1M1.05M1×
  • Output limit128K131K1×
  • +4 more changes
new model
GPT OSS 20B (EU)

via edenai

What changed

First observed in the model catalog

capability
GPT-5.4

via xpersona

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via abacus

What changed

  • temperatureNoYesEnabled
new model
Qwen3.8 27B

via neuralwatt

What changed

First observed in the model catalog

new model
GLM-5.3-Flash

via zhipuai-coding-plan

What changed

First observed in the model catalog

capability
GPT-5.4

via abacus

What changed

  • temperatureNoYesEnabled
repriced
Hermes High

via nano-gpt

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,image,pdftext
  • Context1M1.05M1×
  • Output limit128K131K1×
  • +4 more changes
capability
GPT-5.1

via anyapi

What changed

  • temperatureNoYesEnabled
repriced
Llama-3.3-70B-Instruct

via hyper

What changed

  • Input price$0.598/M$0.638/M▲ 6.7%
  • Output price$0.738/M$0.768/M▲ 4.1%
  • cache write$0.299/M$0.319/M▲ 6.7%
capability
GPT-5.4 Mini

via azure

What changed

  • temperatureNoYesEnabled
new model
Qwen3.8 Flash

via empiriolabs

What changed

First observed in the model catalog

context
Nemotron 3 Nano 30B A3B

via kilo

What changed

  • Output limit236K228K−3%
capability
GPT-5.4

via anyapi

What changed

  • temperatureNoYesEnabled
capability
GPT 5.2

via nano-gpt

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via databricks

What changed

  • temperatureNoYesEnabled
capability
GPT-5.1

via impossibl

What changed

  • temperatureNoYesEnabled
context
Llama 3.1 70B

via wandb

What changed

  • Context128K131K1×
  • Output limit128K131K1×
capability
Grok 4.6

via vercel

What changed

  • structured output—YesEnabled
  • knowledge—2026-02-01
new model
GLM-5.3-Flash

via zai

What changed

First observed in the model catalog

repriced
Claw Medium

via nano-gpt

What changed

  • Context205K1.05M5.1×
  • Input limit205K1.05M5.1×
  • Input price$0.315/M$1.40/M▲ 344.4%
  • Output price$1.26/M$4.40/M▲ 249.2%
  • +1 more changes
capability
GPT-5.4 nano

via databricks

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via opper

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via llmgateway

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via openai

What changed

  • temperatureNoYesEnabled
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.36/M$1.29/M▼ 5.1%
  • Output price$4.40/M$4.22/M▼ 4.1%
  • cache write$0.68/M$0.645/M▼ 5.1%
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output price$0.075/M$0.1/M▲ 33.3%
capability
Seed 1.6 Flash

via volcengine

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
capability
GPT-5.2

via merge-gateway

What changed

  • temperatureNoYesEnabled
new model
Qwen3.8 Flash

via kilo

What changed

First observed in the model catalog

new model
Qwen3.8 2.4T A95B

via hyper

What changed

First observed in the model catalog

capability
GPT-5.4 mini

via crossmodel

What changed

  • temperatureNoYesEnabled
removed
Mistral: Ministral 8B

via kilo

What changed

Removed from the model catalog

new model
GLM-5.3 Flash (Z AI)

via llmgateway-providers

What changed

First observed in the model catalog

capability
GPT-5.4 mini

via requesty

What changed

  • temperatureNoYesEnabled
capability
GPT-5 Chat Latest

via merge-gateway

What changed

  • temperatureYesNoRemoved
context
GLM-5.2

via hyper

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context1.05M1M−5%
context
Qwen3.5 35B-A3B

via kilo

What changed

  • Output limit236K66K−72%
new model
Gemini 3.1 Flash Lite

via edenai

What changed

First observed in the model catalog

repriced
Kimi K2.5

via openrouter

What changed

  • Input price$0.6/M$0.45/M▼ 25.0%
  • Output price$3/M$2.25/M▼ 25.0%
  • Cache read$0.1/M$0.07/M▼ 30.0%
capability
GPT-5.3 Codex

via github-copilot

What changed

  • temperatureNoYesEnabled
context
GLM 5.2

via wandb

What changed

  • Context262K1.05M4×
  • Output limit262K1.05M4×
capability
GPT-5.2

via nearai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via impossibl

What changed

  • temperatureNoYesEnabled
capability
GPT 5.4 Mini

via nano-gpt

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via github-copilot

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4

via requesty

What changed

  • temperatureNoYesEnabled
removed
TheDrummer: Rocinante 12B

via kilo

What changed

Removed from the model catalog

capability
GPT-5.4 (OpenAI)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 Mini

via venice

What changed

  • temperatureNoYesEnabled
capability
GPT 5.4

via nano-gpt

What changed

  • temperatureNoYesEnabled
capability
Grok 4.5

via vercel

What changed

  • structured output—YesEnabled
capability
GPT-5.4 mini

via abacus

What changed

  • temperatureNoYesEnabled
capability
GPT 5.3 Codex

via nano-gpt

What changed

  • temperatureNoYesEnabled
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
capability
GPT-5.4 nano

via model-oracle-ai

What changed

  • temperatureNoYesEnabled
new model
Glm 5.3 Flash

via cloudflare-workers-ai

What changed

First observed in the model catalog

capability
GPT-5.4 mini

via nearai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via openai

What changed

  • temperatureNoYesEnabled
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.41/M$0.424/M▲ 3.4%
  • Output price$1.52/M$1.61/M▲ 6.1%
  • cache write$0.205/M$0.212/M▲ 3.4%
capability
GPT-5.4 mini

via model-oracle-ai

What changed

  • temperatureNoYesEnabled
capability
GPT-5.4 nano

via crossmodel

What changed

  • temperatureNoYesEnabled
capability
GPT-5.3 Codex

via pioneer

What changed

  • temperatureNoYesEnabled
capability
GPT-5.2 (OpenAI)

via llmgateway-providers

What changed

  • temperatureNoYesEnabled
new model
Gemini 3.1 Flash Lite (US)

via edenai

What changed

First observed in the model catalog

126 events
context
MiniMax M3

via merge-gateway

What changed

  • Output limit128K512K4×
repriced
GPT-5.6 Sol

via amazon-bedrock

What changed

  • Input price$5.50/M$4.40/M▼ 20.0%
  • Output price$33/M$22/M▼ 33.3%
  • Cache read$0.55/M$0.44/M▼ 20.0%
  • cache write$6.88/M$5.50/M▼ 20.0%
  • +1 more changes
new model
GLM-5.3-Flash (2x usage)

via opencode-go

What changed

First observed in the model catalog

new model
Qwen3.8 2.4T A95B (Max)

via nano-gpt

What changed

First observed in the model catalog

context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
new model
Gemma 4 26B A4B MeroMero

via nano-gpt

What changed

First observed in the model catalog

context
Qwen3.5 397B-A17B

via kilo

What changed

  • Output limit236K66K−72%
removed
GLM 5.1

via wandb

What changed

Removed from the model catalog

capability
gpt-oss-20b

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
new provider
Seed 2.1 Pro

via volcengine

What changed

volcengine began listing this model

repriced
Qwen: Qwen2.5 VL 72B Instruct

via kilo

What changed

  • Context128K32K−75%
  • Output limit115K29K−75%
  • Input price$0.8/M$0.25/M▼ 68.8%
  • Output price$1/M$0.75/M▼ 25.0%
  • +1 more changes
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.176/M$0.352/M▲ 100.0%
  • Output price$0.528/M$1.06/M▲ 100.0%
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
context
MiniMax-M3

via opper

What changed

  • Output limit128K512K4×
repriced
MiMo-V2.5-Pro

via kilo

What changed

  • Input price$0.435/M$1/M▲ 129.9%
  • Output price$0.87/M$3/M▲ 244.8%
  • Cache read$0.0040/M$0.2/M▲ 4900.0%
repriced
GPT OSS 120B

via edenai

What changed

  • Cache read—$0.15/M
  • cache write—$0.15/M
context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
new model
GLM-5.3 Flash

via merge-gateway

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.077/M$0.089/M▲ 15.1%
  • Output price$0.154/M$0.177/M▲ 15.1%
  • Cache read$0.015/M$0.018/M▲ 15.1%
new model
Qwen3.8 27B

via huggingface

What changed

First observed in the model catalog

deprecated
Ox Alpha Free (Unlimited)

via opencode-go

What changed

  • status—deprecated
new model
MiniMax-M2.7

via gmicloud

What changed

First observed in the model catalog

removed
Nemotron 3 Super

via wandb

What changed

Removed from the model catalog

context
MiniMax-M3

via llmgateway

What changed

  • Output limit128K512K4×
repriced
mistral-medium-3.5

via cortecs

What changed

  • Input price$1.67/M$1.39/M▼ 16.6%
  • Output price$5.57/M$7.13/M▲ 28.0%
  • Cache read—$0.139/M
context
MiniMax M3 (Nebius AI)

via llmgateway-providers

What changed

  • Output limit128K512K4×
context
MiniMax M3

via nano-gpt

What changed

  • Context512K1.05M2×
new provider
Seed Character

via volcengine

What changed

volcengine began listing this model

context
MiniMax-M3

via scnet-token-plan

What changed

  • Context512K1.05M2×
  • Output limit128K512K4×
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▲ 0.1%
  • Output price$0.758/M$0.758/M▲ 0.1%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.758/M$0.758/M▲ 0.1%
  • Output price$0.758/M$0.758/M▲ 0.1%
context
MiniMax-M3

via jalapeno

What changed

  • Output limit128K512K4×
context
MiniMax M2 (MiniMax)

via llmgateway-providers

What changed

  • Context197K205K1×
context
MiniMax-M2 Her

via ofox

What changed

  • Context200K66K−67%
  • Output limit131K2K−98%
deprecated
Ox Alpha Free (Unlimited)

via opencode

What changed

  • status—deprecated
context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
capability
gpt-oss-120b

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
context
MiniMax-M2

via ofox

What changed

  • Context197K205K1×
  • Output limit128K131K1×
new model
Gemma 4 26B A4B MeroMero Thinking

via nano-gpt

What changed

First observed in the model catalog

removed
MiniMax M2.5

via wandb

What changed

Removed from the model catalog

new model
MiniMax-M3

via gmicloud

What changed

First observed in the model catalog

new provider
Seed 1.6 Vision

via volcengine

What changed

volcengine began listing this model

new provider
Seed Evolving

via volcengine

What changed

volcengine began listing this model

context
MiniMax M3 (MiniMax)

via llmgateway-providers

What changed

  • Context512K1.05M2×
context
MiniMax-M3

via ofox

What changed

  • Context512K1.05M2×
  • Output limit128K512K4×
context
MiniMax-M3

via zenmux

What changed

  • Context512K1.05M2×
  • Output limit128K512K4×
repriced
Llama-3.3-70B-Instruct

via openrouter

What changed

  • Output limit16K115K7×
  • Input price$0.1/M$0.71/M▲ 610.0%
  • Output price$0.32/M$0.71/M▲ 121.9%
  • Cache read—$0.71/M
context
MiniMax-M3

via minimax-cn-coding-plan

What changed

  • Context1M1.05M1×
  • Output limit128K512K4×
repriced
Gemma 4 31B IT

via kilo

What changed

  • Output limit236K16K−93%
  • Input price$0.08/M$0.09/M▲ 12.5%
  • Output price$0.35/M$0.34/M▼ 2.9%
  • Cache read$0.01/M$0.05/M▲ 400.0%
context
MiniMax-M2

via minimax-cn

What changed

  • Context197K205K1×
  • Output limit128K131K1×
context
MiniMax-M2

via edenai

What changed

  • Output limit128K131K1×
new model
GLM 5.3 Flash

via nano-gpt

What changed

First observed in the model catalog

context
MiniMax-M2

via huggingface

What changed

  • Output limit128K131K1×
removed
Ox Alpha

via nano-gpt

What changed

Removed from the model catalog

repriced
GPT-5.6 Luna (Global)

via amazon-bedrock

What changed

  • structured outputYesNoRemoved
  • Input price$0.22/M$0.2/M▼ 9.1%
  • Output price$1.32/M$1.20/M▼ 9.1%
  • Cache read$0.022/M$0.02/M▼ 9.1%
  • +2 more changes
repriced
DeepSeek V4 Pro 0813

via edenai

What changed

  • Input price$0.581/M$1.12/M▲ 93.2%
  • Output price$1.74/M$3.37/M▲ 93.2%
repriced
Gemma 4 31B IT

via openrouter

What changed

  • Output limit236K16K−93%
  • Input price$0.1/M$0.09/M▼ 10.0%
  • Cache read$0.1/M$0.05/M▼ 50.0%
repriced
Gemini 3.7 Flash

via venice

What changed

  • Input price$1.88/M$0.938/M▼ 50.0%
  • Output price$9.38/M$4.69/M▼ 50.0%
  • Cache read$0.188/M$0.094/M▼ 50.0%
capability
GPT OSS Safeguard 120B

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▲ 0.1%
  • Output price$1.05/M$1.05/M▲ 0.1%
context
MiniMax-M3

via deepinfra

What changed

  • Output limit128K512K4×
new model
GLM-5.3 Highspeed

via zai-coding-plan

What changed

First observed in the model catalog

new provider
Seed 2.0 Pro

via volcengine

What changed

volcengine began listing this model

repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.04/M$0.06/M▲ 50.0%
  • Output price$0.08/M$0.12/M▲ 50.0%
  • Cache read$0.0080/M$0.012/M▲ 50.0%
context
MiniMax-M2

via minimax-cn-coding-plan

What changed

  • Context197K205K1×
  • Output limit128K131K1×
context
MiniMax-M3

via wafer.ai

What changed

  • Output limit128K512K4×
context
MiniMax-M3

via huggingface

What changed

  • Output limit128K512K4×
context
MiniMax-M2

via minimax

What changed

  • Context197K205K1×
  • Output limit128K131K1×
repriced
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new provider
Qwen3.8 27B

via llmtech

What changed

llmtech began listing this model

context
MiniMax-M3

via minimax-coding-plan

What changed

  • Context1M1.05M1×
  • Output limit128K512K4×
repriced
Hy3

via kilo

What changed

  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • Cache read$0.033/M$0.035/M▲ 6.1%
context
MiniMax-M2

via minimax-coding-plan

What changed

  • Context197K205K1×
  • Output limit128K131K1×
removed
DeepSeek V4 Flash 0731

via edenai

What changed

Removed from the model catalog

new model
GLM-5.3-Flash

via edenai

What changed

First observed in the model catalog

repriced
GPT-5.6 Sol (Global)

via amazon-bedrock

What changed

  • structured outputYesNoRemoved
  • Input price$5.50/M$4/M▼ 27.3%
  • Output price$33/M$20/M▼ 39.4%
  • Cache read$0.55/M$0.4/M▼ 27.3%
  • +2 more changes
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.466/M$0.467/M▲ 0.1%
  • Output price$0.933/M$0.934/M▲ 0.1%
repriced
Qwen2.5 VL 72B Instruct

via openrouter

What changed

  • Output limit115K29K−75%
  • Input price$0.8/M$0.25/M▼ 68.8%
  • Output price$1/M$0.75/M▼ 25.0%
  • Cache read$0.4/M—
context
MiniMax-M3

via hyper

What changed

  • Context512K1.05M2×
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.556/M$0.79/M▲ 42.2%
  • Output price$1.11/M$1.58/M▲ 42.2%
  • Cache read$0.046/M$0.066/M▲ 42.2%
new provider
GPT-4.1 mini

via aixy

What changed

aixy began listing this model

removed
Ox Alpha (retires Aug 26)

via kilo

What changed

Removed from the model catalog

repriced
Qwen3.5 397B-A17B

via openrouter

What changed

  • Output limit236K66K−72%
  • Input price$0.5/M$0.39/M▼ 22.0%
  • Output price$3.60/M$2.34/M▼ 35.0%
  • Cache read$0.3/M—
repriced
Qwen3.6 27B

via openrouter

What changed

  • Output limit82K236K2.9×
  • Input price$0.32/M$0.6/M▲ 87.5%
  • Output price$3.20/M$3.60/M▲ 12.5%
  • Cache read—$0.12/M
new model
GLM-5.3-Flash

via kilo

What changed

First observed in the model catalog

context
MiniMax-M3

via minimax-cn

What changed

  • Context1M1.05M1×
  • Output limit128K512K4×
repriced
GPT-5.6 Terra (Global)

via amazon-bedrock

What changed

  • structured outputYesNoRemoved
  • Input price$2.20/M$2/M▼ 9.1%
  • Output price$13.20/M$12/M▼ 9.1%
  • Cache read$0.22/M$0.2/M▼ 9.1%
  • +2 more changes
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Input price$0.035/M$0.04/M▲ 14.3%
  • Output price$0.28/M$0.08/M▼ 71.4%
new provider
Seed 1.6

via volcengine

What changed

volcengine began listing this model

new model
GLM-5.2

via mistral

What changed

First observed in the model catalog

repriced
Claude Sonnet 5

via venice

What changed

  • Input price$3/M$2/M▼ 33.3%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.3/M$0.2/M▼ 33.3%
  • cache write$3.75/M$2.50/M▼ 33.3%
context
MiniMax-M2

via llmgateway

What changed

  • Context197K205K1×
  • Output limit128K131K1×
context
Meta: Llama 4 Scout

via kilo

What changed

  • Context328K131K−60%
  • Output limit16K8K−50%
capability
GPT OSS Safeguard 20B

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
new provider
Seed 2.0 Code

via volcengine

What changed

volcengine began listing this model

repriced
Gemini 3.6 Flash

via venice

What changed

  • Input price$1.88/M$0.938/M▼ 50.0%
  • Output price$9.38/M$4.69/M▼ 50.0%
  • Cache read$0.188/M$0.094/M▼ 50.0%
context
MiniMax M3 Thinking

via nano-gpt

What changed

  • Context512K1.05M2×
capability
Ox Alpha (retires Aug 26)

via kilo

What changed

  • nameOx AlphaOx Alpha (retires Aug 26)
context
MiniMax-M3

via cline-pass

What changed

  • Context512K1.05M2×
  • Output limit128K512K4×
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Input price$0.035/M$0.04/M▲ 14.3%
  • Output price$0.36/M$0.08/M▼ 77.8%
new provider
Seed 1.8

via volcengine

What changed

volcengine began listing this model

repriced
GPT OSS 20B

via edenai

What changed

  • Cache read—$0.07/M
  • cache write—$0.07/M
context
MiniMax-M3

via gmicloud

What changed

  • Output limit128K512K4×
context
DeepSeek V4 Flash 0731

via edenai

What changed

  • Context1M786K−21%
context
Llama-3.3-70B-Instruct

via kilo

What changed

  • Context131K128K−2%
  • Output limit16K115K7×
removed
Qwen3 Coder 480B A35B

via wandb

What changed

Removed from the model catalog

context
Qwen3.6 27B

via kilo

What changed

  • Output limit82K236K2.9×
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new model
GLM-5.3-Flash

via openrouter

What changed

First observed in the model catalog

repriced
Llama 4 Scout

via openrouter

What changed

  • Output limit16K8K−50%
  • Input price$0.1/M$0.11/M▲ 10.0%
  • Output price$0.3/M$0.34/M▲ 13.3%
  • Cache read—$0.055/M
removed
Ox Alpha

via openrouter

What changed

Removed from the model catalog

new provider
Seed 2.1 Turbo

via volcengine

What changed

volcengine began listing this model

repriced
Kimi K2.6

via inceptron

What changed

  • Input price$0.57/M$0.56/M▼ 1.8%
capability
gpt-oss-120b

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
capability
Inkling

via edenai

What changed

  • structured outputNoYesEnabled
new model
GLM-5.3 Highspeed

via zhipuai-coding-plan

What changed

First observed in the model catalog

repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▲ 0.1%
  • Output price$0.7/M$0.7/M▲ 0.1%
repriced
GPT-5.6 Sol

via merge-gateway

What changed

  • Input price$4/M$5/M▲ 25.0%
  • Output price$24/M$30/M▲ 25.0%
capability
gpt-oss-20b

via amazon-bedrock

What changed

  • open weightsNoYesEnabled
context
MiniMax-M3

via requesty

What changed

  • Output limit128K512K4×
new provider
Seed 2.0 Mini

via volcengine

What changed

volcengine began listing this model

context
MiniMax-M3

via minimax

What changed

  • Context1M1.05M1×
  • Output limit128K512K4×
context
MiniMax-M3

via edenai

What changed

  • Output limit128K512K4×
new provider
Seed 1.6 Flash

via volcengine

What changed

volcengine began listing this model

new provider
Seed 2.0 Lite

via volcengine

What changed

volcengine began listing this model

context
MiniMax-M3

via kenari

What changed

  • Context512K1.05M2×
  • Output limit128K512K4×

390 events
context
Dots3-Note Preview (free)

via openrouter

What changed

  • Output limit512K461K−10%
new model
Qwen 3.8 27B Uncensored

via nano-gpt

What changed

First observed in the model catalog

repriced
GPT-5.6 Sol

via openai

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
  • +1 more changes
context
Qwen3.6 35B-A3B

via openrouter

What changed

  • Output limit262K236K−10%
new model
Qwen 3.8 27B Obliterated

via nano-gpt

What changed

First observed in the model catalog

context
Qwen: Qwen2.5 7B Instruct

via kilo

What changed

  • Output limit33K29K−10%
removed
GPT OSS 120B

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
MiniMax M3 (Free)

via vercel

What changed

First observed in the model catalog

context
Nex AGI: Nex-N2-Pro

via kilo

What changed

  • Output limit262K236K−10%
new model
Gemma 4 26B A4B Uncensored Thinking

via nano-gpt

What changed

First observed in the model catalog

context
IBM: Granite 4.1 8B

via kilo

What changed

  • Output limit131K118K−10%
context
Inkling (free)

via openrouter

What changed

  • Context262K1.05M4×
capability
GPT OSS 120B

via edenai

What changed

  • tool callYesNoRemoved
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.33/M$1.36/M▲ 2.1%
  • Output price$4.31/M$4.40/M▲ 2.0%
  • cache write$0.666/M$0.68/M▲ 2.1%
removed
Gemma 3n 4B

via openrouter

What changed

Removed from the model catalog

context
Llama 3.2 1B Instruct

via openrouter

What changed

  • Output limit60K54K−10%
context
Kimi K2.7 Code

via kilo

What changed

  • Output limit262K236K−10%
removed
Llama 3.2 1B Instruct

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
MoonshotAI Kimi Latest

via kilo

What changed

  • Output limit1.05M944K−10%
context
Qwen2.5 Coder 32B Instruct

via kilo

What changed

  • Output limit33K29K−10%
context
Qwen3 Coder Next

via kilo

What changed

  • Output limit262K236K−10%
context
Trinity Large Thinking

via openrouter

What changed

  • Output limit262K236K−10%
context
GPT OSS 20B

via openrouter

What changed

  • Output limit131K118K−10%
removed
GPT-3.5-turbo

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
Qwen3.5 397B-A17B

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
GLM-5.2

via kilo

What changed

  • Output limit131K262K2×
context
Meta: Muse Spark 1.2 Contributor

via kilo

What changed

  • Output limit1.05M944K−10%
context
Grok 4.5

via kilo

What changed

  • Output limit500K450K−10%
capability
Claude Opus 4.7

via cloudflare-ai-gateway

What changed

  • release date2026-04-142026-04-16
context
Grok 4.5

via openrouter

What changed

  • Output limit500K450K−10%
repriced
Ornith 1.5 35B

via nano-gpt

What changed

  • Cache read$0.01/M$0.05/M▲ 400.0%
context
Grok Latest

via openrouter

What changed

  • Output limit1M450K−55%
context
Hy3 preview

via kilo

What changed

  • Output limit262K236K−10%
context
Mistral: Mixtral 8x22B Instruct

via kilo

What changed

  • Output limit13K52K4×
repriced
DeepSeek V4 Pro 0813

via edenai

What changed

  • Input price$0.627/M$1.12/M▲ 78.9%
  • Output price$1.88/M$3.37/M▲ 78.9%
context
MiniMax-01

via openrouter

What changed

  • Output limit1.00M900K−10%
capability
Llama-3.3-70B-Instruct

via edenai

What changed

  • tool callYesNoRemoved
context
Mistral: Ministral 3 8B 2512

via kilo

What changed

  • Output limit33K210K6.4×
context
OpenAI: GPT-3.5 Turbo Instruct

via kilo

What changed

  • Output limit4K4K−10%
repriced
GPT-5.6

via openai

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
  • +1 more changes
context
MythoMax 13B

via openrouter

What changed

  • Output limit4K4K−10%
context
Mistral: Saba

via kilo

What changed

  • Output limit33K26K−20%
context
Mistral Medium 3

via openrouter

What changed

  • Output limit131K105K−20%
context
Hermes 4 70B

via openrouter

What changed

  • Output limit131K118K−10%
new model
Qwen3.6 35B-A3B

via evroc

What changed

First observed in the model catalog

context
Ministral 3 3B 2512

via openrouter

What changed

  • Output limit131K105K−20%
removed
Gemma Sea Lion V4 27B It

via cloudflare-ai-gateway

What changed

Removed from the model catalog

repriced
GPT-5.4

via cloudflare-ai-gateway

What changed

  • Context1.05M1M−5%
  • tiers[object Object]
removed
Llama 4 Scout 17B 16E Instruct

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
MiniMax M2.7 (free)

via openrouter

What changed

First observed in the model catalog

context
Grok Build 0.1

via openrouter

What changed

  • Output limit256K230K−10%
repriced
Grok 4.5

via opencode

What changed

  • Cache read$0.5/M$0.3/M▼ 40.0%
  • tiers[object Object][object Object]
removed
Mistral Small 3.1 24B Instruct

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
Qwen3 Max

via cloudflare-ai-gateway

What changed

First observed in the model catalog

capability
Qwen3-Next 80B-A3B Instruct

via merge-gateway

What changed

  • nameQwen3 Next 80B A3B InstructQwen3-Next 80B-A3B Instruct
context
Mistral Large 3

via openrouter

What changed

  • Output limit262K210K−20%
new provider
DeepSeek V4 Flash

via pendra

What changed

pendra began listing this model

context
GPT-5 Nano

via cloudflare-ai-gateway

What changed

  • Context400K128K−68%
removed
GPT-5.3 Codex

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
Qwen3 Coder Next (EU)

via edenai

What changed

First observed in the model catalog

context
Nous: Hermes 4 70B

via kilo

What changed

  • Output limit131K118K−10%
context
Skyfall 36B V2

via openrouter

What changed

  • Output limit33K29K−10%
repriced
Ornith 1.5 35B Thinking

via nano-gpt

What changed

  • Cache read$0.01/M$0.05/M▲ 400.0%
repriced
GLM-5.2

via openrouter

What changed

  • Output limit131K262K2×
  • Input price$0.966/M$1.19/M▲ 23.2%
  • Output price$3.04/M$3.74/M▲ 23.2%
  • Cache read$0.193/M$0.221/M▲ 14.4%
context
Mistral Small 3.1 24B

via openrouter

What changed

  • Output limit128K102K−20%
new provider
Agnes 2.5 Flash

via agnes

What changed

agnes began listing this model

new model
Ornith 1.5 397B Thinking

via nano-gpt

What changed

First observed in the model catalog

context
Qwen3 235B A22B Thinking 2507

via openrouter

What changed

  • Output limit33K118K3.6×
context
Nano Banana 2 Lite

via openrouter

What changed

  • Output limit66K59K−10%
context
TheDrummer: Cydonia 24B V4.1

via kilo

What changed

  • Output limit131K118K−10%
context
Nemotron 3 Nano 30B A3B

via kilo

What changed

  • Output limit262K236K−10%
context
Mistral: Mistral Small 3.1 24B

via kilo

What changed

  • Output limit128K102K−20%
new provider
GLM-4.7-Flash

via pendra

What changed

pendra began listing this model

context
Mistral: Ministral 3 14B 2512

via kilo

What changed

  • Output limit52K210K4×
context
Mistral Large 3

via kilo

What changed

  • Output limit52K210K4×
context
OpenAI: GPT-3.5 Turbo (older v0613)

via kilo

What changed

  • Output limit4K4K−10%
context
DeepSeek V4 Pro 0813

via kilo

What changed

  • Output limit384K944K2.5×
repriced
GPT-5.5

via cloudflare-ai-gateway

What changed

  • Context1.05M1M−5%
  • Cache read$0.5/M—
  • tiers[object Object]
new provider
Agnes 2.0 Flash

via agnes

What changed

agnes began listing this model

context
Qwen3.5 397B-A17B

via kilo

What changed

  • Output limit262K236K−10%
context
AllenAI: Olmo 3 32B Think

via kilo

What changed

  • Output limit66K59K−10%
context
GPT-5.4 nano

via cloudflare-ai-gateway

What changed

  • Context400K128K−68%
context
ByteDance Seed: Seed 2.1 Turbo

via kilo

What changed

  • Output limit262K236K−10%
context
Sonar Deep Research

via openrouter

What changed

  • Output limit128K115K−10%
context
Gemma 4 26B A4B Uncensored

via nano-gpt

What changed

  • Context131K262K2×
  • Input limit131K262K2×
context
Sonar Reasoning Pro

via openrouter

What changed

  • Output limit128K115K−10%
repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
removed
o1-pro

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Kimi K2.6

via openrouter

What changed

  • Output limit262K236K−10%
context
Qwen3.5 122B-A10B

via openrouter

What changed

  • Output limit262K236K−10%
context
Nex-N2-Pro

via openrouter

What changed

  • Output limit262K236K−10%
context
GPT-5 Mini

via cloudflare-ai-gateway

What changed

  • Context400K128K−68%
removed
Google: Gemma 3n 4B

via kilo

What changed

Removed from the model catalog

removed
Kimi K2.6

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Muse Spark 1.2

via openrouter

What changed

  • Output limit1.05M944K−10%
new provider
NeoSmith NeoLite

via neosmith

What changed

neosmith began listing this model

context
Mistral: Voxtral Small 24B 2507

via kilo

What changed

  • Output limit6K26K4×
context
Meta: Llama 3.2 3B Instruct

via kilo

What changed

  • Output limit131K118K−10%
context
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

  • Context66K262K4×
  • Input limit66K262K4×
context
Nemotron 3 Super (free)

via openrouter

What changed

  • Output limit262K236K−10%
repriced
Devstral 2

via edenai

What changed

  • Input price$0.4/M$0.44/M▲ 10.0%
  • Output price$2/M$2.20/M▲ 10.0%
  • Cache read—$0.044/M
context
MiniMax-M2.7

via openrouter

What changed

  • Output limit131K177K1.3×
context
Qwen3.5 27B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
context
GLM-5.3

via llmgateway

What changed

  • Context1M1.05M1×
new model
Grok 4.20 (Reasoning)

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Qwen3 Coder Next

via openrouter

What changed

  • Output limit262K236K−10%
context
Qwen2.5 Coder 32B Instruct

via openrouter

What changed

  • Output limit33K29K−10%
new model
MiniMax: MiniMax M2.7 (free)

via kilo

What changed

First observed in the model catalog

repriced
GLM-5.1

via openrouter

What changed

  • Output limit128K203K1.6×
  • Input price$0.966/M$1.26/M▲ 30.4%
  • Output price$3.04/M$3.96/M▲ 30.4%
  • Cache read$0.179/M$0.234/M▲ 30.4%
context
GLM-5.1

via kilo

What changed

  • Context200K203K1×
  • Output limit128K203K1.6×
new model
Grok 4.6

via opencode-go

What changed

First observed in the model catalog

removed
Nemotron 3 Super 120B

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

  • Context66K262K4×
  • Input limit66K262K4×
context
TheDrummer: Rocinante 12B

via kilo

What changed

  • Output limit66K59K−10%
context
Mixtral 8x22B Instruct

via openrouter

What changed

  • Output limit66K52K−20%
context
Muse Glimmer 30B

via kilo

What changed

  • Output limit131K118K−10%
context
Kimi K3

via kilo

What changed

  • Output limit1.05M944K−10%
context
Step 3.7 Flash

via openrouter

What changed

  • Output limit256K230K−10%
context
Llama-3.1-8B-Instruct

via kilo

What changed

  • Output limit131K118K−10%
removed
GPT OSS 20B

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Grok 4.3

via openrouter

What changed

  • Output limit1M900K−10%
removed
Qwen3.6 35B-A3B

via evroc

What changed

Removed from the model catalog

context
Nano Banana 2 Lite

via kilo

What changed

  • Output limit66K59K−10%
context
Qwen3.6 35B-A3B

via kilo

What changed

  • Output limit262K236K−10%
removed
GPT-5.2 Chat

via cloudflare-ai-gateway

What changed

Removed from the model catalog

removed
Llama 3.3 70B Instruct fp8 Fast

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Saba

via openrouter

What changed

  • Output limit33K26K−20%
removed
GPT-5.6

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Llama Guard 4 12B

via openrouter

What changed

  • Context1.05M164K−84%
new provider
NeoSmith Pro

via neosmith

What changed

neosmith began listing this model

new provider
GPT OSS 120B

via pendra

What changed

pendra began listing this model

repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output price$0.12/M$0.075/M▼ 37.5%
  • Cache read$0.01/M$0.0080/M▼ 20.0%
context
Kimi K3

via openrouter

What changed

  • Output limit1.05M944K−10%
context
GLM-5.1

via openrouter

What changed

  • Output limit203K182K−10%
context
Trinity Large Thinking

via kilo

What changed

  • Output limit262K236K−10%
removed
Llama Guard 3 8B

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
MiniMax: MiniMax M2.7 (free)

via kilo

What changed

  • Output limit197K177K−10%
context
Nemotron 3 Ultra 550B A55B

via openrouter

What changed

  • Output limit16K461K28.1×
context
R1 Distill Llama 70B

via openrouter

What changed

  • Output limit8K7K−10%
new provider
Standard Compute

via standardcompute

What changed

standardcompute began listing this model

context
Muse Glimmer 30B

via openrouter

What changed

  • Output limit131K118K−10%
context
GPT OSS 120B

via openrouter

What changed

  • Output limit131K118K−10%
context
AionLabs: Aion-RP 1.0 (8B)

via kilo

What changed

  • Output limit33K29K−10%
context
MiniMax M3 (free)

via openrouter

What changed

  • Output limit1.05M944K−10%
removed
GPT-5.2

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Devstral 2

via kilo

What changed

  • Output limit262K210K−20%
context
Qwen3-Next 80B-A3B Instruct

via kilo

What changed

  • Output limit262K236K−10%
repriced
GLM-5

via hyper

What changed

  • Input price$0.86/M$0.91/M▲ 5.8%
  • Output price$2.78/M$2.81/M▲ 1.1%
  • cache write$0.43/M$0.455/M▲ 5.8%
context
Nous: Hermes 4 405B

via kilo

What changed

  • Output limit26K118K4.5×
context
Mistral: Mistral Medium 3.5

via kilo

What changed

  • Output limit262K210K−20%
new model
Qwen3.7 Plus

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Grok Build 0.1

via kilo

What changed

  • Output limit256K230K−10%
removed
GPT-5.2 Pro

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
TheDrummer: Skyfall 36B V2

via kilo

What changed

  • Output limit33K29K−10%
context
Nemotron 3 Nano 30B A3B

via openrouter

What changed

  • Output limit262K236K−10%
context
IBM: Granite 4.0 Micro

via kilo

What changed

  • Output limit131K118K−10%
removed
GLM-4.7-Flash

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Qwen3.5 9B

via kilo

What changed

  • Output limit262K236K−10%
context
Aion-RP 1.0 (8B)

via openrouter

What changed

  • Output limit33K29K−10%
context
Seed 2.1 Turbo

via openrouter

What changed

  • Output limit262K236K−10%
removed
Llama 3.2 3B Instruct

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
DeepSeek V4 Pro

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Qwen3-Coder 30B-A3B Instruct

via openrouter

What changed

  • Output limit262K236K−10%
context
Llama 3 8B Lunaris

via openrouter

What changed

  • Output limit16K7K−55%
context
Kimi K2.6

via kilo

What changed

  • Output limit262K236K−10%
capability
Claude Sonnet 5

via cloudflare-ai-gateway

What changed

  • release date2026-06-292026-06-30
repriced
Inkling

via openrouter

What changed

  • Input price$1/M$0.95/M▼ 5.0%
  • Cache read$0.17/M$0.16/M▼ 5.9%
removed
Qwen2.5 Coder 32B Instruct

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
GPT-3.5 Turbo Instruct

via openrouter

What changed

  • Output limit4K4K−10%
new model
GLM-5.3

via cline-pass

What changed

First observed in the model catalog

context
Muse Spark 1.2 Contributor

via openrouter

What changed

  • Output limit1.05M944K−10%
new provider
NeoSmith Basic

via neosmith

What changed

neosmith began listing this model

new model
Qwen3.7 Max

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Dots Studio: Dots3-Note Preview (free)

via kilo

What changed

  • Output limit512K461K−10%
context
MiniMax: MiniMax-01

via kilo

What changed

  • Output limit1.00M900K−10%
context
Upstage: Solar Pro 3

via kilo

What changed

  • Output limit131K118K−10%
context
Hy3 preview

via openrouter

What changed

  • Output limit262K236K−10%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.467/M$0.466/M▼ 0.0%
  • Output price$0.933/M$0.933/M▼ 0.0%
removed
Nemotron 3.5 Lightning 30B (Free)

via vercel

What changed

Removed from the model catalog

new provider
Llama-3.3-70B-Instruct

via pendra

What changed

pendra began listing this model

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.056/M$0.089/M▲ 58.2%
  • Output price$0.112/M$0.177/M▲ 58.2%
  • Cache read$0.011/M$0.018/M▲ 58.2%
new model
MiniMax: MiniMax M3 (free)

via kilo

What changed

First observed in the model catalog

context
Olmo 3 32B Think

via openrouter

What changed

  • Output limit66K59K−10%
context
Ministral 3 14B 2512

via openrouter

What changed

  • Output limit262K210K−20%
context
Ornith 1.5 9B

via nano-gpt

What changed

  • Context131K262K2×
  • Input limit131K262K2×
removed
Llama 3.1 8B Instruct fp8

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Hermes 4 405B

via openrouter

What changed

  • Output limit131K118K−10%
context
DeepSeek V3.2

via openrouter

What changed

  • Output limit164K147K−10%
context
Gemma 4 31B MeroMero v2 Thinking

via nano-gpt

What changed

  • Context66K262K4×
  • Input limit66K262K4×
context
Mistral Large

via kilo

What changed

  • Output limit26K102K4×
context
Qwen: Qwen3 235B A22B Thinking 2507

via kilo

What changed

  • Output limit262K118K−55%
removed
Kimi K2.7 Code

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
DeepSeek: DeepSeek V3 0324

via kilo

What changed

  • Output limit164K147K−10%
repriced
OpenAI GPT-5.6 Terra

via digitalocean

What changed

  • Input price$1/M$2/M▲ 100.0%
  • Output price$6/M$12/M▲ 100.0%
  • Cache read$0.1/M$0.2/M▲ 100.0%
  • tiers[object Object][object Object]
context
SpaceXAI: Grok 4.20

via kilo

What changed

  • Output limit2M1.80M−10%
context
Mistral: Codestral 2508

via kilo

What changed

  • Output limit51K205K4×
context
Perplexity: Sonar Reasoning Pro

via kilo

What changed

  • Output limit26K115K4.5×
context
Sao10K: Llama 3 8B Lunaris

via kilo

What changed

  • Output limit16K7K−55%
new model
Ornith 1.5 9B Thinking

via nano-gpt

What changed

First observed in the model catalog

context
Mistral: Ministral 8B

via kilo

What changed

  • Output limit128K102K−20%
context
Mistral Large 2407

via openrouter

What changed

  • Output limit131K105K−20%
context
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Output limit384K944K2.5×
new provider
Qwen3.8 27B

via iteracompute

What changed

iteracompute began listing this model

context
Solar Pro 3

via openrouter

What changed

  • Output limit131K118K−10%
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.178/M$0.188/M▲ 5.6%
  • Output price$0.68/M$0.7/M▲ 2.9%
  • cache write$0.089/M$0.094/M▲ 5.6%
context
Mistral Medium 3.5

via openrouter

What changed

  • Output limit262K210K−20%
context
Qwen3.5 9B

via openrouter

What changed

  • Output limit262K236K−10%
context
Gemma 3 27B

via openrouter

What changed

  • Output limit131K118K−10%
context
Muse Spark 1.1

via kilo

What changed

  • Output limit1.05M944K−10%
context
Ornith 1.5 9B Thinking

via nano-gpt

What changed

  • Context131K262K2×
  • Input limit131K262K2×
capability
Qwen3-Next 80B-A3B (Thinking)

via merge-gateway

What changed

  • nameQwen3 Next 80B A3B ThinkingQwen3-Next 80B-A3B (Thinking)
capability
Gemma 4 26B A4B Uncensored

via nano-gpt

What changed

  • reasoningNoYesEnabled
removed
GPT-5.3 Chat (latest)

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Perplexity: Sonar

via kilo

What changed

  • Output limit25K114K4.5×
new model
GLM-5.3 (Baidu)

via llmgateway-providers

What changed

First observed in the model catalog

context
DeepSeek V4 Flash 0731

via kilo

What changed

  • Output limit131K944K7.2×
removed
DeepSeek Reasoner

via deepseek

What changed

Removed from the model catalog

repriced
Qwen3.8 27B

via openrouter

What changed

  • Input price$0.4/M$0.425/M▲ 6.2%
  • Output price$3/M$2.55/M▼ 15.0%
  • Cache read$0.05/M$0.085/M▲ 70.0%
  • cache write—$0.531/M
repriced
GPT-5.6 Luna

via cloudflare-ai-gateway

What changed

  • tiers[object Object]
context
Ministral 8B

via openrouter

What changed

  • Output limit128K102K−20%
context
Gemma 4 26B A4B Uncensored Thinking

via nano-gpt

What changed

  • Context131K262K2×
  • Input limit131K262K2×
capability
Claude Fable 5

via cloudflare-ai-gateway

What changed

  • release date2026-06-072026-06-09
repriced
GLM 5.2

via inceptron

What changed

  • Output price$2.90/M$2.40/M▼ 17.2%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.0%
  • Output price$0.758/M$0.758/M▼ 0.0%
context
GPT-4.1 nano

via cloudflare-ai-gateway

What changed

  • statusdeprecated—
  • Context1.05M1M−5%
capability
o4-mini

via cloudflare-ai-gateway

What changed

  • statusdeprecated—
removed
Qwen3 30B A3b fp8

via cloudflare-ai-gateway

What changed

Removed from the model catalog

capability
DeepSeek V4 Flash 0731

via edenai

What changed

  • tool callYesNoRemoved
context
Grok 4.6

via openrouter

What changed

  • Output limit500K450K−10%
context
Nano Banana Pro

via kilo

What changed

  • Context66K131K2×
context
DeepSeek V3 0324

via openrouter

What changed

  • Output limit164K147K−10%
context
GPT-5.4 mini

via cloudflare-ai-gateway

What changed

  • Context400K128K−68%
context
Cydonia 24B V4.1

via openrouter

What changed

  • Output limit131K118K−10%
new provider
Qwen3-Coder 30B-A3B Instruct

via pendra

What changed

pendra began listing this model

capability
o3-mini

via cloudflare-ai-gateway

What changed

  • statusdeprecated—
new model
Grok 4.3

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Mistral Large

via openrouter

What changed

  • Output limit128K102K−20%
repriced
GPT-5.5 Pro

via cloudflare-ai-gateway

What changed

  • Context1.05M1M−5%
  • tiers[object Object]
context
MoonshotAI Kimi Latest

via openrouter

What changed

  • Output limit1.05M944K−10%
context
Qwen3.8 27B

via kilo

What changed

  • Context262K1M3.8×
removed
o3-pro

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit1.05M944K−10%
context
GPT-5

via cloudflare-ai-gateway

What changed

  • Context400K128K−68%
removed
Qwen: Qwen Plus 0728 (thinking)

via kilo

What changed

Removed from the model catalog

context
Mistral: Ministral 3 3B 2512

via kilo

What changed

  • Output limit33K105K3.2×
context
Llama 3.2 3B Instruct

via openrouter

What changed

  • Output limit131K118K−10%
context
DeepSeek: DeepSeek V3.1

via kilo

What changed

  • Output limit161K145K−10%
repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Input price$0.066/M$0.14/M▲ 112.8%
  • Output price$0.132/M$0.28/M▲ 112.8%
  • Cache read$0.013/M$0.028/M▲ 112.8%
context
Kimi K2.5

via kilo

What changed

  • Output limit262K236K−10%
new model
Qwen 3.8 27B Obliterated Thinking

via nano-gpt

What changed

First observed in the model catalog

context
Perplexity: Sonar Deep Research

via kilo

What changed

  • Output limit26K115K4.5×
context
Ministral 3 8B 2512

via openrouter

What changed

  • Output limit262K210K−20%
repriced
OpenAI GPT-5.6 Luna

via digitalocean

What changed

  • Input price$0.1/M$0.2/M▲ 100.0%
  • Output price$0.6/M$1.20/M▲ 100.0%
  • Cache read$0.01/M$0.02/M▲ 100.0%
  • tiers[object Object][object Object]
removed
DeepSeek Chat

via deepseek

What changed

Removed from the model catalog

context
Rocinante 12B

via openrouter

What changed

  • Output limit66K59K−10%
context
DeepSeek V4 Flash Latest

via kilo

What changed

  • Output limit1.05M944K−10%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.176/M$0.352/M▲ 100.0%
  • Output price$0.528/M$1.06/M▲ 100.0%
repriced
GPT-5.6 Terra

via cloudflare-ai-gateway

What changed

  • tiers[object Object]
context
Nano Banana 2

via openrouter

What changed

  • Output limit66K59K−10%
capability
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

  • reasoningNoYesEnabled
removed
MiniMax M3 (free)

via openrouter

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output price$0.12/M$0.075/M▼ 37.5%
  • Cache read$0.01/M$0.0080/M▼ 20.0%
repriced
Reka Flash 3

via kilo

What changed

  • Cache read—$0.1/M
new model
Grok 4.20 (Non-Reasoning)

via cloudflare-ai-gateway

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.522/M$0.79/M▲ 51.4%
  • Output price$1.04/M$1.58/M▲ 51.4%
  • Cache read$0.043/M$0.066/M▲ 51.4%
removed
o1

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Google: Gemma 3 27B

via kilo

What changed

  • Output limit131K118K−10%
context
Qwen3.5 35B-A3B

via openrouter

What changed

  • Output limit262K236K−10%
new model
Ornith 1.5 397B

via nano-gpt

What changed

First observed in the model catalog

new model
Grok 4.6

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Llama-3.1-8B-Instruct

via openrouter

What changed

  • Output limit131K118K−10%
context
Muse Spark 1.2

via kilo

What changed

  • Output limit1.05M944K−10%
context
Qwen3.5 397B-A17B

via openrouter

What changed

  • Output limit262K236K−10%
repriced
Hy3

via kilo

What changed

  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • Cache read$0.033/M$0.035/M▲ 6.1%
new model
MiniMax M3 (free)

via openrouter

What changed

First observed in the model catalog

new model
Gemma 4 31B MeroMero v2 Thinking

via nano-gpt

What changed

First observed in the model catalog

context
TheDrummer: UnslopNemo 12B

via kilo

What changed

  • Output limit33K26K−20%
new model
Qwen 3.8 27B Uncensored Thinking

via nano-gpt

What changed

First observed in the model catalog

context
Nex-N2-Mini

via openrouter

What changed

  • Output limit262K236K−10%
new model
Qwen3.8 Max

via cline-pass

What changed

First observed in the model catalog

repriced
MoonshotAI Kimi Latest

via openrouter

What changed

  • Output limit975K1.05M1.1×
  • Input price$2.60/M$2.80/M▲ 7.7%
  • Output price$13/M$14/M▲ 7.7%
context
Gemma 4 31B IT

via kilo

What changed

  • Output limit262K236K−10%
new provider
Qwen3.6 27B

via pendra

What changed

pendra began listing this model

context
UnslopNemo 12B

via openrouter

What changed

  • Output limit33K26K−20%
context
Qwen3.5 397B A17B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
context
DeepSeek: R1 Distill Llama 70B

via kilo

What changed

  • Output limit8K7K−10%
context
Nemotron 3 Ultra 550B A55B

via kilo

What changed

  • Output limit512K461K−10%
removed
Qwen 3.8 27B Uncensored

via nano-gpt

What changed

Removed from the model catalog

context
Qwen3-Coder 30B-A3B Instruct

via kilo

What changed

  • Output limit262K236K−10%
context
Qwen3.5 35B-A3B

via kilo

What changed

  • Output limit262K236K−10%
repriced
OpenAI GPT-5.6 Sol

via digitalocean

What changed

  • Input price$2/M$5/M▲ 150.0%
  • Output price$10/M$30/M▲ 200.0%
  • Cache read$0.2/M$0.5/M▲ 150.0%
  • tiers[object Object][object Object]
context
Kimi K2.7 Code

via openrouter

What changed

  • Output limit262K236K−10%
deprecated
Grok 4.5

via opencode-go

What changed

  • status—deprecated
  • Cache read$0.5/M$0.3/M▼ 40.0%
  • tiers[object Object][object Object]
new model
Gemini 3.7 Flash

via vivgrid

What changed

First observed in the model catalog

new model
Wan v3.0 Video

via vercel

What changed

First observed in the model catalog

context
GPT OSS 20B

via kilo

What changed

  • Output limit131K118K−10%
context
MiniMax M2.7 (free)

via openrouter

What changed

  • Output limit197K177K−10%
repriced
Reka Flash 3

via openrouter

What changed

  • Cache read—$0.1/M
new model
Qwen3.8-27B

via evroc

What changed

First observed in the model catalog

context
Grok 4.20 Multi-Agent

via openrouter

What changed

  • Output limit2M1.80M−10%
context
Devstral 2

via openrouter

What changed

  • Output limit262K210K−20%
repriced
GPT-5.6 Sol

via cloudflare-ai-gateway

What changed

  • Input price$5/M$2/M▼ 60.0%
  • Output price$30/M$10/M▼ 66.7%
  • Cache read$0.5/M$0.25/M▼ 50.0%
  • cache write$6.25/M$3.13/M▼ 50.0%
  • +1 more changes
removed
GPT-4

via cloudflare-ai-gateway

What changed

Removed from the model catalog

removed
Deepseek R1 Distill Qwen 32B

via cloudflare-ai-gateway

What changed

Removed from the model catalog

removed
Llama 3.2 11B Vision Instruct

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
MiniMax-M2.7

via kilo

What changed

  • Output limit131K177K1.3×
context
Codestral 2508

via openrouter

What changed

  • Output limit256K205K−20%
context
SpaceXAI: Grok 4.20 Multi-Agent

via kilo

What changed

  • Output limit2M1.80M−10%
context
Kimi K2.5

via openrouter

What changed

  • Output limit262K236K−10%
context
Granite 4.1 8B

via openrouter

What changed

  • Output limit131K118K−10%
context
Mistral Small 4

via kilo

What changed

  • Output limit262K210K−20%
repriced
GPT OSS 20B

via kilo

What changed

  • Input price$0.03/M$0.02/M▼ 33.3%
  • Output price$0.13/M$0.1/M▼ 23.1%
  • Cache read$0.03/M—
context
Qwen2.5 VL 72B Instruct

via openrouter

What changed

  • Output limit128K115K−10%
context
Gemma 4 31B IT

via openrouter

What changed

  • Output limit262K236K−10%
new model
Qwen3.8 Max

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
DeepSeek V3.2

via kilo

What changed

  • Output limit164K147K−10%
repriced
DeepSeek V4 Flash 0731

via merge-gateway

What changed

  • Input price$0.035/M$0.22/M▲ 528.6%
  • Output price$0.07/M$0.66/M▲ 842.9%
new model
Granite 4.2 8B

via wandb

What changed

First observed in the model catalog

repriced
Reka Edge

via kilo

What changed

  • Cache read—$0.1/M
repriced
DeepSeek V4 Pro (Baidu)

via llmgateway-providers

What changed

  • Input price$1.69/M$1.32/M▼ 21.9%
  • Output price$3.38/M$3.96/M▲ 17.2%
  • Cache read$0.14/M$0.132/M▼ 5.7%
context
Mistral Large 2407

via kilo

What changed

  • Output limit33K105K3.2×
removed
Glm 5.2

via cloudflare-ai-gateway

What changed

Removed from the model catalog

new model
Qwen 3.6 35B A3B Uncensored Thinking

via nano-gpt

What changed

First observed in the model catalog

removed
GPT-5.3 Codex Spark

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Tencent: Hunyuan A13B Instruct

via kilo

What changed

  • Output limit131K118K−10%
repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.11/M$0.12/M▲ 9.1%
  • Output price$0.408/M$0.42/M▲ 2.9%
  • cache write$0.055/M$0.06/M▲ 9.1%
context
Mistral: Mistral Medium 3.1

via kilo

What changed

  • Output limit26K105K4×
context
Hunyuan A13B Instruct

via openrouter

What changed

  • Output limit131K118K−10%
context
Nex AGI: Nex-N2-Mini

via kilo

What changed

  • Output limit262K236K−10%
new model
GLM-5.3

via zai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash (Baidu)

via llmgateway-providers

What changed

  • Input price$0.14/M$0.44/M▲ 214.3%
  • Output price$0.28/M$1.32/M▲ 371.4%
  • Cache read$0.028/M$0.044/M▲ 57.1%
context
Grok 4.20

via openrouter

What changed

  • Output limit2M1.80M−10%
context
MythoMax 13B

via kilo

What changed

  • Output limit4K4K−10%
context
Nano Banana 2

via kilo

What changed

  • Output limit66K59K−10%
context
Inkling Small (free)

via openrouter

What changed

  • Context262K1.05M4×
context
Grok 4.3

via kilo

What changed

  • Output limit4K900K219.7×
repriced
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
context
NVIDIA: Nemotron 3 Super (free)

via kilo

What changed

  • Output limit262K236K−10%
repriced
GPT-5.4 Pro

via cloudflare-ai-gateway

What changed

  • Context1.05M1M−5%
  • tiers[object Object]
context
MiniMax: MiniMax M3 (free)

via kilo

What changed

  • Output limit1.05M944K−10%
new model
Kimi K3

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Qwen3-Next 80B-A3B Instruct

via openrouter

What changed

  • Output limit262K236K−10%
context
xAI: Grok Latest

via kilo

What changed

  • Output limit500K450K−10%
context
GPT OSS 120B

via kilo

What changed

  • Output limit131K118K−10%
context
Grok 4.6

via kilo

What changed

  • Output limit500K450K−10%
removed
Gemma 4 26B A4B IT

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
GPT-5.1

via cloudflare-ai-gateway

What changed

  • Context400K128K−68%
context
Qwen3.5 122B-A10B

via kilo

What changed

  • Output limit262K236K−10%
repriced
Nemotron 3.5 Lightning 30B

via vercel

What changed

  • Context1M262K−74%
  • Output limit33K131K4×
  • Input price$0/M$0.05/M
  • Output price$0/M$0.2/M
  • +1 more changes
context
Mistral: Mistral Medium 3

via kilo

What changed

  • Output limit26K105K4×
context
Muse Spark 1.1

via openrouter

What changed

  • Output limit1.05M944K−10%
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new model
Qwen3 Max (EU)

via edenai

What changed

First observed in the model catalog

context
Google: Gemma 3n 4B

via kilo

What changed

  • Output limit7K29K4.5×
removed
GPT-4 Turbo

via cloudflare-ai-gateway

What changed

Removed from the model catalog

context
Voxtral Small 24B 2507

via openrouter

What changed

  • Output limit32K26K−20%
new model
Agnes 2.5 Pro Alpha

via agnes

What changed

First observed in the model catalog

removed
Qwq 32B

via cloudflare-ai-gateway

What changed

Removed from the model catalog

repriced
gpt-oss-safeguard-20b

via vercel

What changed

  • Context131K128K−2%
  • Output limit66K16K−76%
  • Input limit66K112K1.7×
  • Input price$0.075/M$0.07/M▼ 6.7%
  • +2 more changes
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.0%
  • Output price$0.7/M$0.7/M▼ 0.0%
new model
GPT OSS Safeguard 120B

via vercel

What changed

First observed in the model catalog

context
Qwen: Qwen2.5 VL 72B Instruct

via kilo

What changed

  • Output limit128K115K−10%
context
Granite 4.0 Micro

via openrouter

What changed

  • Output limit131K118K−10%
context
DeepSeek V3.1

via openrouter

What changed

  • Output limit161K145K−10%
context
Mistral Small 4

via openrouter

What changed

  • Output limit262K210K−20%
context
Sonar

via openrouter

What changed

  • Output limit127K114K−10%
removed
DeepSeek Reasoner

via edenai

What changed

Removed from the model catalog

context
Qwen2.5 7B Instruct

via openrouter

What changed

  • Output limit33K29K−10%
context
Step 3.7 Flash

via kilo

What changed

  • Output limit256K230K−10%
context
Gemma 4 31B MeroMero v2

via nano-gpt

What changed

  • Context66K262K4×
  • Input limit66K262K4×
removed
Granite 4.0 H Micro

via cloudflare-ai-gateway

What changed

Removed from the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▼ 0.0%
  • Output price$1.05/M$1.05/M▼ 0.0%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.758/M$0.758/M▼ 0.0%
  • Output price$0.758/M$0.758/M▼ 0.0%
capability
Claude Opus 4.6

via cloudflare-ai-gateway

What changed

  • release date2026-02-042026-02-05
context
GLM 5.2 (free)

via openrouter

What changed

  • Output limit256K230K−10%
new model
Minimax M2.7 (Free)

via vercel

What changed

First observed in the model catalog

context
Meta: Llama 3.2 1B Instruct

via kilo

What changed

  • Output limit60K54K−10%
context
Gemma 3n 4B

via openrouter

What changed

  • Output limit33K29K−10%
context
Phi 4

via openrouter

What changed

  • Output limit16K15K−10%
context
Mistral Medium 3.1

via openrouter

What changed

  • Output limit262K105K−60%
context
GPT-3.5 Turbo (older v0613)

via openrouter

What changed

  • Output limit4K4K−10%
new provider
NeoSmith Maestro

via neosmith

What changed

neosmith began listing this model

repriced
MoonshotAI Kimi Latest

via kilo

What changed

  • Context975K1.05M1.1×
  • Output limit975K1.05M1.1×
  • Input price$2.60/M$2.80/M▲ 7.7%
  • Output price$13/M$14/M▲ 7.7%
removed
GPT-5 Pro

via cloudflare-ai-gateway

What changed

Removed from the model catalog

repriced
Reka Edge

via openrouter

What changed

  • Cache read—$0.1/M
context
Microsoft: Phi 4

via kilo

What changed

  • Output limit16K15K−10%
removed
Qwen Plus 0728 (thinking)

via openrouter

What changed

Removed from the model catalog

repriced
Kimi K2.6

via inceptron

What changed

  • Output price$3.41/M$3.39/M▼ 0.6%
new model
Grok 4.5

via cloudflare-ai-gateway

What changed

First observed in the model catalog

context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%

102 events
context
ReMM SLERP 13B

via openrouter

What changed

  • Output limit6K4K−33%
repriced
Qwen Plus

via alibaba-cn

What changed

  • Cache read—$0.012/M
  • cache write—$0.144/M
repriced
MiniMax-M2.7

via openrouter

What changed

  • Input price$0.24/M$0.3/M▲ 25.0%
  • Output price$0.96/M$1.20/M▲ 25.0%
  • Cache read$0.048/M$0.06/M▲ 25.0%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.468/M$0.467/M▼ 0.3%
  • Output price$0.936/M$0.933/M▼ 0.3%
repriced
GLM-5.1

via hyper

What changed

  • Input price$1.36/M$1.33/M▼ 2.1%
  • Output price$4.40/M$4.31/M▼ 2.0%
  • cache write$0.68/M$0.666/M▼ 2.1%
context
UnslopNemo 12B

via openrouter

What changed

  • Output limit1.02M33K−97%
repriced
Kimi K2.5

via openrouter

What changed

  • Input price$0.45/M$0.6/M▲ 33.3%
  • Output price$2.25/M$3/M▲ 33.3%
  • Cache read$0.07/M$0.1/M▲ 42.9%
repriced
Mistral: Mistral Small 3.2 24B

via kilo

What changed

  • Input price$0.075/M$0.11/M▲ 46.7%
  • Output price$0.2/M$0.33/M▲ 65.0%
  • Cache read—$0.011/M
context
GPT OSS 120B

via hyper

What changed

  • Context131K128K−2%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.176/M$0.352/M▲ 100.0%
  • Output price$0.528/M$1.06/M▲ 100.0%
repriced
GPT Latest

via nano-gpt

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
repriced
Kimi K2.6

via inceptron

What changed

  • Input price$0.6/M$0.57/M▼ 5.0%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▼ 0.3%
  • Output price$1.05/M$1.05/M▼ 0.3%
capability
Qwen3.8 27B

via runinfra

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
repriced
DeepSeek V4 Pro 0813

via edenai

What changed

  • Input price$0.627/M$1.16/M▲ 85.3%
  • Output price$1.88/M$3.48/M▲ 85.3%
repriced
Qwen3.5 397B-A17B

via openrouter

What changed

  • Output limit66K262K4×
  • Input price$0.39/M$0.5/M▲ 28.2%
  • Output price$2.34/M$3.60/M▲ 53.8%
  • Cache read—$0.3/M
removed
Nemotron Nano 9B v2 (US)

via edenai

What changed

Removed from the model catalog

repriced
Qwen3 30B A3B

via openrouter

What changed

  • Output limit8K16K2×
  • Input price$0.13/M$0.12/M▼ 7.7%
  • Output price$0.52/M$0.5/M▼ 3.8%
context
Qwen3 30B A3B

via kilo

What changed

  • Context131K41K−69%
  • Output limit8K16K2×
new model
Claude Opus 5 (Azure Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
GPT Chat Latest

via nano-gpt

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
repriced
Qwen3.8 27B

via kilo

What changed

  • Input price$0.5/M$0.425/M▼ 15.0%
  • Output price$3/M$2.55/M▼ 15.0%
  • Cache read$0.1/M$0.085/M▼ 15.0%
  • cache write$0.625/M$0.531/M▼ 15.0%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.76/M$0.758/M▼ 0.3%
  • Output price$0.76/M$0.758/M▼ 0.3%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.3%
  • Output price$0.702/M$0.7/M▼ 0.3%
removed
Trinity Mini

via vercel

What changed

Removed from the model catalog

removed
o3-deep-research

via vercel

What changed

Removed from the model catalog

context
Kimi K2.7 Code

via hyper

What changed

  • Context256K262K1×
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Input price$0.04/M$0.035/M▼ 12.5%
  • Output price$0.08/M$0.13/M▲ 62.5%
  • Cache read$0.0080/M$0.01/M▲ 25.0%
repriced
Qwen3.6 27B

via openrouter

What changed

  • Output limit262K82K−69%
  • Input price$0.6/M$0.32/M▼ 46.7%
  • Output price$3.60/M$3.20/M▼ 11.1%
  • Cache read$0.12/M—
new model
Claude Opus 4.7 (Azure Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

context
Qwen3.5 122B-A10B

via kilo

What changed

  • Output limit66K262K4×
repriced
OpenAI GPT-5.6 Sol

via digitalocean

What changed

  • Input price$5/M$2/M▼ 60.0%
  • Output price$30/M$10/M▼ 66.7%
  • Cache read$0.5/M$0.2/M▼ 60.0%
  • tiers[object Object][object Object]
removed
GPT 4o Mini Search Preview

via vercel

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Output limit384K131K−66%
  • Input price$0.08/M$0.14/M▲ 75.0%
  • Output price$0.18/M$0.28/M▲ 55.6%
  • Cache read$0.016/M$0.028/M▲ 75.0%
repriced
Mistral Nemo

via kilo

What changed

  • Input price$0.019/M$0.165/M▲ 768.4%
  • Output price$0.03/M$0.165/M▲ 450.0%
  • Cache read—$0.017/M
context
Thinking Machines: Inkling Small (free)

via kilo

What changed

  • Context262K1.05M4×
context
Qwen3.6 Plus

via llmgateway

What changed

  • Context262K1M3.8×
repriced
Qwen3 Coder Plus (Alibaba Cloud)

via llmgateway-providers

What changed

  • Input price$6/M$1/M▼ 83.3%
  • Output price$60/M$5/M▼ 91.7%
  • Cache read$1.20/M$0.2/M▼ 83.3%
  • cache write$7.50/M$1.25/M▼ 83.3%
repriced
Kimi K2.6

via openrouter

What changed

  • Input price$0.541/M$0.95/M▲ 75.4%
  • Output price$2.28/M$4/M▲ 75.4%
  • Cache read$0.091/M$0.16/M▲ 75.4%
removed
Ling 2.6 1T

via nano-gpt

What changed

Removed from the model catalog

new provider
Claude Opus 4.8

via agentrouter

What changed

agentrouter began listing this model

new model
Qwen3.8 27B TEE

via nano-gpt

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via hyper

What changed

  • Input price$0.479/M$0.44/M▼ 8.2%
  • Output price$1.44/M$1.32/M▼ 8.2%
  • Cache read$0.015/M$0.044/M▲ 188.7%
new model
Devstral 2

via openrouter

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via vercel

What changed

  • Input price$1.32/M$0.66/M▼ 50.0%
  • Output price$3.96/M$1.98/M▼ 50.0%
  • Cache read$0.132/M$0.066/M▼ 50.0%
removed
Nemotron Nano 9B v2

via edenai

What changed

Removed from the model catalog

new model
DeepSeek V4 Flash 0731

via edenai

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.049/M$0.059/M▲ 20.1%
  • Output price$0.098/M$0.117/M▲ 20.1%
  • Cache read$0.0098/M$0.012/M▲ 20.1%
new model
Claude Opus 4.8 (Azure Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Qwen3 Coder Plus

via llmgateway

What changed

  • Input price$6/M$1/M▼ 83.3%
  • Output price$60/M$5/M▼ 91.7%
  • Cache read$1.20/M$0.2/M▼ 83.3%
  • cache write$7.50/M$1.25/M▼ 83.3%
context
ReMM SLERP 13B

via kilo

What changed

  • Output limit6K4K−33%
context
Qwen3 235B A22B Thinking 2507

via openrouter

What changed

  • Context262K131K−50%
new provider
Claude Opus 5

via agentrouter

What changed

agentrouter began listing this model

removed
Ling 2.6 Flash

via nano-gpt

What changed

Removed from the model catalog

context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
new model
Seedance 2.0 Mini

via vercel

What changed

First observed in the model catalog

context
Thinking Machines: Inkling (free)

via kilo

What changed

  • Context262K1.05M4×
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▼ 0.3%
  • Output price$0.76/M$0.758/M▼ 0.3%
repriced
Llama-3.3-70B-Instruct

via hyper

What changed

  • Input price$0.638/M$0.598/M▼ 6.3%
  • Output price$0.768/M$0.738/M▼ 3.9%
  • cache write$0.319/M$0.299/M▼ 6.3%
removed
Ling-2.6-flash

via openrouter

What changed

Removed from the model catalog

removed
Nemotron Nano 9B V2 (free)

via openrouter

What changed

Removed from the model catalog

new model
LongCat-2.0

via opencode-go

What changed

First observed in the model catalog

repriced
Qwen3 Max (Alibaba Cloud)

via llmgateway-providers

What changed

  • Input price$3/M$1.20/M▼ 60.0%
  • Output price$15/M$6/M▼ 60.0%
  • Cache read$0.6/M$0.24/M▼ 60.0%
  • cache write$3.75/M$1.50/M▼ 60.0%
repriced
GPT OSS 120B

via hyper

What changed

  • Input price$0.188/M$0.178/M▼ 5.3%
  • Output price$0.7/M$0.68/M▼ 2.9%
  • cache write$0.094/M$0.089/M▼ 5.3%
repriced
Kimi K2.7 Code

via openrouter

What changed

  • Cache read$0.17/M$0.19/M▲ 11.8%
new model
Claude Fable 5 (Azure Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
GPT 5.6 Sol Pro

via nano-gpt

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
new model
Claude Sonnet 5 (Azure Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

removed
Ling-2.6-1T

via openrouter

What changed

Removed from the model catalog

removed
Ring 2.6 1T

via nano-gpt

What changed

Removed from the model catalog

repriced
Kimi K2.7 Code

via inceptron

What changed

  • Cache read$0.17/M$0.19/M▲ 11.8%
context
DeepSeek V4 Flash 0731

via edenai

What changed

  • Context1M786K−21%
context
Qwen3.5 397B-A17B

via kilo

What changed

  • Output limit66K262K4×
context
DeepSeek V4 Flash 0731

via kilo

What changed

  • Output limit384K131K−66%
repriced
Qwen3.6 Plus (Alibaba Cloud)

via llmgateway-providers

What changed

  • Context262K1M3.8×
  • cache write—$0.625/M
new provider
GPT-5.6 Sol

via agentrouter

What changed

agentrouter began listing this model

context
Qwen3.6 27B

via kilo

What changed

  • Output limit262K82K−69%
removed
Magistral Small

via vercel

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash 0731 TEE

via chutes

What changed

  • Input price$0.14/M$0.44/M▲ 214.3%
  • Output price$0.28/M$1.32/M▲ 371.4%
  • Cache read$0.014/M$0.044/M▲ 214.3%
removed
inclusionAI: Ling-2.6-flash (retires Aug 24)

via kilo

What changed

Removed from the model catalog

repriced
OpenAI GPT-5.6 Luna

via digitalocean

What changed

  • Input price$0.2/M$0.1/M▼ 50.0%
  • Output price$1.20/M$0.6/M▼ 50.0%
  • Cache read$0.02/M$0.01/M▼ 50.0%
  • tiers[object Object][object Object]
repriced
MiniMax-M2.7

via hyper

What changed

  • Input price$0.408/M$0.41/M▲ 0.5%
  • Output price$1.51/M$1.52/M▲ 0.5%
  • cache write$0.204/M$0.205/M▲ 0.5%
repriced
Qwen3.6 35B A3B (Alibaba Cloud)

via llmgateway-providers

What changed

  • Input price$0.248/M$0.375/M▲ 51.2%
  • Output price$1.49/M$2.25/M▲ 51.5%
removed
inclusionAI: Ring-2.6-1T (retires Aug 24)

via kilo

What changed

Removed from the model catalog

repriced
Hy3

via kilo

What changed

  • Input price$0.132/M$0.14/M▲ 6.1%
  • Output price$0.528/M$0.58/M▲ 9.8%
  • Cache read$0.033/M$0.035/M▲ 6.1%
removed
Ring-2.6-1T

via openrouter

What changed

Removed from the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.397/M$0.519/M▲ 30.8%
  • Output price$0.794/M$1.04/M▲ 30.8%
  • Cache read$0.033/M$0.043/M▲ 30.8%
new model
Devstral 2

via kilo

What changed

First observed in the model catalog

removed
Nemotron 3 Nano 30B A3B (free)

via openrouter

What changed

Removed from the model catalog

context
Qwen3.5 122B-A10B

via openrouter

What changed

  • Output limit66K262K4×
repriced
GLM-5

via hyper

What changed

  • Input price$0.91/M$0.86/M▼ 5.5%
  • Output price$2.93/M$2.78/M▼ 5.1%
  • cache write$0.455/M$0.43/M▼ 5.5%
repriced
Inkling

via openrouter

What changed

  • Input price$0.95/M$1/M▲ 5.3%
  • Cache read$0.16/M$0.17/M▲ 6.3%
new model
Claude Opus 4.6 (Azure Anthropic)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Input price$0.04/M$0.035/M▼ 12.5%
  • Output price$0.08/M$0.13/M▲ 62.5%
  • Cache read$0.0080/M$0.01/M▲ 25.0%
repriced
GPT 5.6 Sol

via nano-gpt

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
context
TheDrummer: UnslopNemo 12B

via kilo

What changed

  • Context1.02M33K−97%
  • Output limit1.02M33K−97%
removed
DeepSeek V4 Flash 0731

via edenai

What changed

Removed from the model catalog

repriced
OpenAI GPT-5.6 Terra

via digitalocean

What changed

  • Input price$2/M$1/M▼ 50.0%
  • Output price$12/M$6/M▼ 50.0%
  • Cache read$0.2/M$0.1/M▼ 50.0%
  • tiers[object Object][object Object]
removed
Nemotron Nano 12B 2 VL (free)

via openrouter

What changed

Removed from the model catalog

removed
inclusionAI: Ling-2.6-1T (retires Aug 24)

via kilo

What changed

Removed from the model catalog

removed
Magistral Medium (latest)

via vercel

What changed

Removed from the model catalog

70 events
new provider
Sonar

via opper

What changed

opper began listing this model

repriced
DeepSeek V4 Pro 0813

via edenai

What changed

  • Input price$0.627/M$1.16/M▲ 85.3%
  • Output price$1.88/M$3.48/M▲ 85.3%
new provider
Grok 4.3

via opper

What changed

opper began listing this model

new model
Gemma 4 31B MeroMero v2

via nano-gpt

What changed

First observed in the model catalog

repriced
Kimi K2.6

via openrouter

What changed

  • Input price$0.541/M$0.95/M▲ 75.4%
  • Output price$2.28/M$4/M▲ 75.4%
  • Cache read$0.091/M$0.16/M▲ 75.4%
new provider
Gemini 3 Flash Preview

via opper

What changed

opper began listing this model

new provider
GPT-5.5 Pro

via opper

What changed

opper began listing this model

repriced
Kimi K2.7 Code

via aki-io

What changed

  • Cache read—$0.18/M
new model
DeepSeek V4 Flash 0731

via aki-io

What changed

First observed in the model catalog

capability
Inkling Small

via openrouter

What changed

  • structured outputYesNoRemoved
new provider
Muse Spark 1.2

via opper

What changed

opper began listing this model

repriced
Qwen3.8 27B TEE

via chutes

What changed

  • Input price$0.4/M$0.35/M▼ 12.5%
  • Output price$3/M$2.75/M▼ 8.3%
  • Cache read$0.04/M$0.035/M▼ 12.5%
new provider
GPT-5.3 Chat (latest)

via opper

What changed

opper began listing this model

new model
Qwen3.8 27B

via aki-io

What changed

First observed in the model catalog

context
MiniMax-M2.7

via kilo

What changed

  • Context205K197K−4%
removed
MiniMax-M2.5

via aki-io

What changed

Removed from the model catalog

new provider
Claude Haiku 4.5 (latest)

via opper

What changed

opper began listing this model

repriced
Qwen3 30B A3B

via openrouter

What changed

  • Output limit8K16K2×
  • Input price$0.13/M$0.12/M▼ 7.7%
  • Output price$0.52/M$0.5/M▼ 3.8%
new provider
GPT-5.3 Codex

via opper

What changed

opper began listing this model

new provider
Gemini 3.1 Pro Preview

via opper

What changed

opper began listing this model

repriced
MiniMax-M2.7

via openrouter

What changed

  • Input price$0.3/M$0.24/M▼ 20.0%
  • Output price$1.20/M$0.96/M▼ 20.0%
  • Cache read$0.06/M$0.048/M▼ 20.0%
context
Qwen3.5 122B-A10B

via openrouter

What changed

  • Output limit66K262K4×
new provider
Devstral 2

via opper

What changed

opper began listing this model

new provider
GPT-5.6 Sol

via opper

What changed

opper began listing this model

new provider
Kimi K3

via opper

What changed

opper began listing this model

repriced
Inkling

via openrouter

What changed

  • Input price$0.95/M$1/M▲ 5.3%
  • Cache read$0.16/M$0.17/M▲ 6.3%
repriced
DeepSeek V4 Flash 0731

via merge-gateway

What changed

  • Input price$0.22/M$0.035/M▼ 84.1%
  • Output price$0.66/M$0.07/M▼ 89.4%
context
Qwen3.5 122B-A10B

via kilo

What changed

  • Output limit66K262K4×
repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Input price$0.064/M$0.06/M▼ 6.1%
  • Output price$0.128/M$0.18/M▲ 40.8%
  • Cache read$0.013/M$0.028/M▲ 119.1%
new provider
GPT-5.4

via opper

What changed

opper began listing this model

new provider
Claude Sonnet 5

via opper

What changed

opper began listing this model

new provider
GPT-5.4 mini

via opper

What changed

opper began listing this model

new provider
Claude Opus 5

via opper

What changed

opper began listing this model

new provider
Gemini 3.7 Flash

via opper

What changed

opper began listing this model

new model
Ornith 1.5 9B

via nano-gpt

What changed

First observed in the model catalog

repriced
GPT-5.6 Sol

via requesty

What changed

  • Input price$4.50/M$3.60/M▼ 20.0%
  • Output price$27/M$18/M▼ 33.3%
  • Cache read$0.45/M$0.36/M▼ 20.0%
  • cache write—$4.50/M
  • +1 more changes
new provider
Claude Fable 5

via opper

What changed

opper began listing this model

repriced
Gemma 4 26B A4B IT

via kilo

What changed

  • Input price$0.05/M$0.042/M▼ 16.0%
  • Output price$0.25/M$0.22/M▼ 12.0%
new provider
GPT-5.4 nano

via opper

What changed

opper began listing this model

new provider
Sonar Reasoning Pro

via opper

What changed

opper began listing this model

new provider
Gemini 3.5 Flash

via opper

What changed

opper began listing this model

context
Qwen3 30B A3B

via kilo

What changed

  • Context131K41K−69%
  • Output limit8K16K2×
repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
new provider
Mistral Small 4

via opper

What changed

opper began listing this model

new provider
GPT-5.6 Terra

via opper

What changed

opper began listing this model

new provider
Claude Sonnet 4.5 (latest)

via opper

What changed

opper began listing this model

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Input price$0.414/M$0.397/M▼ 4.1%
  • Output price$0.828/M$0.794/M▼ 4.1%
  • Cache read$0.034/M$0.033/M▼ 4.1%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.057/M$0.056/M▼ 2.4%
  • Output price$0.115/M$0.112/M▼ 2.4%
  • Cache read$0.011/M$0.011/M▼ 2.4%
removed
Amazon Nova Micro 1.0

via nano-gpt

What changed

Removed from the model catalog

new provider
GPT-5.6 Luna

via opper

What changed

opper began listing this model

capability
Inkling Small

via kilo

What changed

  • structured outputYesNoRemoved
capability
DeepSeek V4 Pro 0813

via edenai

What changed

  • tool callNoYesEnabled
  • structured outputNoYesEnabled
new provider
MiniMax-M3

via opper

What changed

opper began listing this model

new provider
Mistral Large 3

via opper

What changed

opper began listing this model

new provider
Gemini 3.7 Flash (EU)

via opper

What changed

opper began listing this model

new provider
GPT-5.4 Pro

via opper

What changed

opper began listing this model

new provider
Grok Build 0.1

via opper

What changed

opper began listing this model

new provider
Claude Opus 4.6

via opper

What changed

opper began listing this model

new provider
Claude Sonnet 4.6

via opper

What changed

opper began listing this model

new provider
Claude Opus 4.7

via opper

What changed

opper began listing this model

new provider
Grok 4.6

via opper

What changed

opper began listing this model

new provider
Gemini 3.5 Flash Lite

via opper

What changed

opper began listing this model

new provider
Grok 4.5

via opper

What changed

opper began listing this model

new provider
Claude Opus 4.5 (latest)

via opper

What changed

opper began listing this model

repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.176/M$0.352/M▲ 100.0%
  • Output price$0.528/M$1.06/M▲ 100.0%
new provider
Claude Opus 4.8

via opper

What changed

opper began listing this model

repriced
Gemini Flash Latest

via nano-gpt

What changed

  • tool callNoYesEnabled
  • structured outputNoYesEnabled
  • modalities.inputtext,image,audiotext,image,video,audio,pdf
  • Context1.05M1.05M−0%
  • +5 more changes
new provider
GPT-5.5

via opper

What changed

opper began listing this model

new provider
Sonar Pro

via opper

What changed

opper began listing this model

repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Input price$0.064/M$0.06/M▼ 6.1%
  • Output price$0.128/M$0.18/M▲ 40.8%
  • Cache read$0.013/M$0.028/M▲ 119.1%

121 events
context
MiniMax-M2.5

via kilo

What changed

  • Context198K200K1×
  • Output limit33K128K3.9×
repriced
Kimi K2.6

via openrouter

What changed

  • Input price$0.58/M$0.56/M▼ 3.3%
  • Output price$2.44/M$2.36/M▼ 3.3%
  • Cache read$0.098/M$0.094/M▼ 3.3%
removed
GLM-4.7-Flash

via crof

What changed

Removed from the model catalog

capability
DeepSeek V4 Pro 0813

via edenai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
capability
DeepSeek V4 Pro 0813

via runinfra

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
removed
GPT-4o Mini Search Preview

via llmgateway

What changed

Removed from the model catalog

repriced
MiniMax-M2.5

via openrouter

What changed

  • Output limit33K128K3.9×
  • Output price$0.95/M$1.08/M▲ 13.7%
  • Cache read$0.03/M$0.027/M▼ 10.0%
repriced
DeepSeek V4 Flash

via edenai

What changed

  • Context1.05M1M−5%
  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
capability
DeepSeek V4 Pro 0813

via kilo

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
DeepSeek V4 Pro 0813

via edenai

What changed

  • Input price$0.603/M$1.16/M▲ 92.8%
  • Output price$1.81/M$3.48/M▲ 92.8%
new model
Thinking Machines: Inkling (free)

via kilo

What changed

First observed in the model catalog

removed
GLM-4.7

via crof

What changed

Removed from the model catalog

new model
Gemma 4 26B A4B Uncensored

via nano-gpt

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro 0813

via requesty

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
capability
DeepSeek V4 Pro 0813

via fireworks-ai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
GPT 5.6 Sol (Fast)

via vercel

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
removed
GLM-5

via crof

What changed

Removed from the model catalog

repriced
Qwen3-Next 80B-A3B Instruct

via openrouter

What changed

  • Output limit16K262K16×
  • Input price$0.09/M$0.1/M▲ 11.1%
  • Cache read—$0.07/M
repriced
DeepSeek V4 Flash 0731

via vercel

What changed

  • Input price$0.13/M$0.076/M▼ 41.5%
  • Output price$0.26/M$0.153/M▼ 41.2%
  • Cache read$0.028/M$0.014/M▼ 50.0%
repriced
DeepSeek V4 Flash 0731

via kilo

What changed

  • Input price$0.14/M$0.44/M▲ 214.3%
  • Output price$0.28/M$1.32/M▲ 371.4%
capability
DeepSeek V4 Pro 0813

via nano-gpt

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
removed
gpt-oss-20b (free)

via openrouter

What changed

Removed from the model catalog

new model
DeepSeek V4 Flash Vision Exp

via crossmodel

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro

via openrouter

What changed

  • Output limit393K384K−2%
  • Input price$1.60/M$0.549/M▼ 65.7%
  • Output price$3.20/M$1.10/M▼ 65.7%
  • Cache read$0.135/M$0.046/M▼ 66.1%
removed
Kimi K2.5

via crof

What changed

Removed from the model catalog

new model
Meta: Muse Spark 1.2 Contributor

via kilo

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro 0813 Thinking

via nano-gpt

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
capability
DeepSeek V4 Pro 0813

via edenai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
DeepSeek V3.1

via openrouter

What changed

  • Output limit33K161K4.9×
  • Input price$0.25/M$0.55/M▲ 120.0%
  • Output price$0.95/M$1.65/M▲ 73.7%
  • Cache read$0.13/M$0.55/M▲ 323.1%
new model
Hy-MT2-7B

via openrouter

What changed

First observed in the model catalog

new model
Claude Fable 5

via google-vertex

What changed

First observed in the model catalog

removed
GPT-4o Mini Search Preview (OpenAI)

via llmgateway-providers

What changed

Removed from the model catalog

context
Kimi K3

via hyper

What changed

  • Output limit131K16K−88%
repriced
Qwen3.8 27B

via openrouter

What changed

  • Input price$0.45/M$0.4/M▼ 11.1%
  • Output price$3.20/M$3/M▼ 6.3%
new model
Muse Spark 1.2 Contributor

via openrouter

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
context
Qwen3.8 2.4T A95B

via kilo

What changed

  • Context1M1.05M1×
  • Output limit262K131K−50%
capability
DeepSeek V4 Pro 0813

via huggingface

What changed

  • last updated2026-08-122026-08-22
repriced
Muse Glimmer 30B

via openrouter

What changed

  • Input price$0.3/M$0.35/M▲ 16.7%
  • Output price$1.10/M$1.50/M▲ 36.4%
capability
DeepSeek V4 Pro

via umans-ai-coding-plan

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
GPT OSS 120B

via openrouter

What changed

  • Input price$0.03/M$0.037/M▲ 23.3%
  • Cache read$0.03/M—
repriced
DeepSeek V4 Flash Latest

via openrouter

What changed

  • Output limit1.05M262K−75%
  • Input price$0.065/M$0.064/M▼ 1.7%
  • Output price$0.18/M$0.128/M▼ 29.0%
  • Cache read$0.02/M$0.013/M▼ 36.1%
repriced
Mistral Small 3.2 24B

via openrouter

What changed

  • Context256K131K−49%
  • Input price$0.094/M$0.075/M▼ 20.0%
  • Output price$0.25/M$0.2/M▼ 20.0%
capability
DeepSeek V4 Pro 0813

via vercel

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
capability
DeepSeek V4 Pro 0813

via merge-gateway

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
context
Qwen3.5 122B A10B

via merge-gateway

What changed

  • attachmentYesNoRemoved
  • modalities.inputtext,imagetext
  • Context256K131K−49%
  • Output limit64K33K−49%
context
MiMo-V2.5-Pro

via llmgateway

What changed

  • Context1M1.05M1×
context
DeepSeek: DeepSeek V3.1

via kilo

What changed

  • Context164K161K−2%
  • Output limit33K161K4.9×
repriced
GPT-5.6 Sol Pro

via openrouter

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
capability
DeepSeek V4 Pro 0813

via togetherai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
OpenAI GPT Latest

via openrouter

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
new model
MiMo V2.5 Pro (DeepInfra)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Kimi K3

via nvidia

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via deepseek

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via nano-gpt

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Latest

via kilo

What changed

  • Context1.05M262K−75%
  • Output limit1.05M262K−75%
  • Input price$0.065/M$0.064/M▼ 1.7%
  • Output price$0.18/M$0.128/M▼ 29.0%
  • +1 more changes
context
Qwen3.8 2.4T A95B

via openrouter

What changed

  • Output limit262K131K−50%
repriced
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
removed
GPT-4o Search Preview (OpenAI)

via llmgateway-providers

What changed

Removed from the model catalog

new model
Grok 4.6 (Vertex AI (OpenAI-compatible))

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • Input price$0.22/M$0.44/M▲ 100.0%
  • Output price$0.66/M$1.32/M▲ 100.0%
  • Cache read$0.0070/M$0.014/M▲ 100.0%
repriced
DeepSeek V3.2

via openrouter

What changed

  • Output limit66K164K2.5×
  • Input price$0.269/M$0.26/M▼ 3.3%
  • Output price$0.4/M$0.38/M▼ 5.0%
  • Cache read$0.135/M$0.13/M▼ 3.3%
capability
DeepSeek V4 Pro 0813

via ofox

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.14/M$0.22/M▲ 57.1%
  • Output price$0.28/M$0.66/M▲ 135.7%
  • Cache read$0.028/M$0.0070/M▼ 75.0%
repriced
DeepSeek V4 Pro 0813

via openrouter

What changed

  • Input price$1.19/M$1.12/M▼ 5.6%
  • Output price$3.56/M$3.37/M▼ 5.6%
  • Cache read$0.04/M$0.037/M▼ 5.6%
capability
DeepSeek V4 Pro 0813

via edenai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
capability
DeepSeek V4 Pro 0813

via openrouter

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
GPT-5.6 Sol

via edenai

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
capability
DeepSeek V4 Pro 0813

via baseten

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
new model
MiMo V2.5 (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

new model
Inkling Small (free)

via openrouter

What changed

First observed in the model catalog

repriced
GPT-5.6 Sol

via edenai

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
context
Qwen3.5 35B A3B

via merge-gateway

What changed

  • attachmentNoYesEnabled
  • modalities.inputtexttext,image
  • Context131K256K2×
  • Output limit33K64K2×
capability
DeepSeek V4 Pro 0813

via hyper

What changed

  • open weightsNoYesEnabled
capability
DeepSeek V4 Pro 0813

via arcee

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
capability
DeepSeek V4 Pro 0813

via digitalocean

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
DeepSeek V4 Pro

via edenai

What changed

  • Context1.05M1M−5%
  • Input price$0.66/M$1.32/M▲ 100.0%
  • Output price$1.98/M$3.96/M▲ 100.0%
  • Cache read$0.022/M$0.044/M▲ 100.0%
repriced
Nemotron 3.5 Lightning 30B

via vercel

What changed

  • Context262K1M3.8×
  • Output limit131K33K−75%
  • Input price$0.05/M$0/M▼ 100.0%
  • Output price$0.2/M$0/M▼ 100.0%
  • +1 more changes
repriced
Mistral: Mistral Small 3.2 24B

via kilo

What changed

  • Context256K128K−50%
  • Input price$0.094/M$0.075/M▼ 20.0%
  • Output price$0.25/M$0.2/M▼ 20.0%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.081/M$0.078/M▼ 3.5%
  • Output price$0.162/M$0.157/M▼ 3.5%
  • Cache read$0.016/M$0.016/M▼ 3.5%
repriced
GPT-5.6 Sol

via merge-gateway

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$24/M▼ 20.0%
capability
DeepSeek V4 Pro 0813

via alibaba-token-plan

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
GPT-5.6 Sol

via kilo

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
new model
Nemotron 3.5 Lightning 30B (Free)

via vercel

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro

via umans-ai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
GPT-5.6 Sol (50% Off)

via opencode

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
removed
MiniMax-M2.5

via crof

What changed

Removed from the model catalog

new model
Thinking Machines: Inkling Small (free)

via kilo

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.201/M$0.352/M▲ 75.2%
  • Output price$0.603/M$1.06/M▲ 75.2%
capability
DeepSeek V4 Pro 0813

via alibaba-token-plan-cn

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
removed
Kimi K2.5 (Lightning)

via crof

What changed

Removed from the model catalog

new model
MiMo V2.5 (DeepInfra)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
Qwen3.8 27B

via kilo

What changed

  • Input price$0.575/M$0.5/M▼ 13.0%
  • Output price$3.45/M$3/M▼ 13.0%
  • Cache read$0.115/M$0.1/M▼ 13.0%
  • cache write$0.719/M$0.625/M▼ 13.0%
capability
DeepSeek V4 Pro 0813

via deepinfra

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
new model
Claude Fable 5

via google-vertex-anthropic

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro

via deepseek

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
context
MiMo-V2.5

via llmgateway

What changed

  • Context1M1.05M1×
new model
Tencent: Hy-MT2-7B

via kilo

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro 0813

via empiriolabs

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
new model
DeepSeek V4 Flash 0731

via edenai

What changed

First observed in the model catalog

new model
Inkling (free)

via openrouter

What changed

First observed in the model catalog

new model
Ox Alpha

via nano-gpt

What changed

First observed in the model catalog

removed
Cogito v2.1 671B

via openrouter

What changed

Removed from the model catalog

context
Qwen3.8-Max

via digitalocean

What changed

  • Context262K1M3.8×
repriced
GPT-5.6 Sol

via openrouter

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
context
Qwen3-Next 80B-A3B Instruct

via kilo

What changed

  • Output limit16K262K16×
repriced
GPT 5.6 Sol

via vercel

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
new model
DeepSeek V4 Flash 0731

via ofox

What changed

First observed in the model catalog

capability
DeepSeek V4 Pro 0813

via cloudflare-workers-ai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
new model
DeepSeek V4 Flash 0731

via nvidia

What changed

First observed in the model catalog

new model
Qwen3.8 27B

via crof

What changed

First observed in the model catalog

repriced
Venice Uncensored

via nano-gpt

What changed

  • Output limit16K8K−50%
  • Output price$0.4/M$1.80/M▲ 350.0%
  • Cache read$0.2/M$0.4/M▲ 100.0%
new model
MiMo V2.5 Pro (NovitaAI)

via llmgateway-providers

What changed

First observed in the model catalog

removed
GPT-4o Search Preview

via llmgateway

What changed

Removed from the model catalog

repriced
GPT-5.6 Sol

via kilo

What changed

  • Input price$5/M$4/M▼ 20.0%
  • Output price$30/M$20/M▼ 33.3%
  • Cache read$0.5/M$0.4/M▼ 20.0%
  • cache write$6.25/M$5/M▼ 20.0%
removed
Deep Cogito: Cogito v2.1 671B

via kilo

What changed

Removed from the model catalog

capability
DeepSeek V4 Pro 0813

via edenai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
repriced
OpenAI GPT Latest

via kilo

What changed

  • Input price$2.50/M$2/M▼ 20.0%
  • Output price$15/M$10/M▼ 33.3%
  • Cache read$0.25/M$0.2/M▼ 20.0%
  • cache write$3.13/M$2.50/M▼ 20.0%
context
DeepSeek V3.2

via kilo

What changed

  • Output limit66K164K2.5×
capability
DeepSeek V4 Pro 0813

via edenai

What changed

  • open weightsNoYesEnabled
  • last updated2026-08-122026-08-22
context
DeepSeek V4 Pro

via kilo

What changed

  • Context1.05M1.02M−2%
  • Output limit393K384K−2%

42 events
new model
DeepSeek: DeepSeek V4 Flash Vision Exp

via kilo

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via vercel

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.402/M$0.201/M▼ 50.0%
  • Output price$1.21/M$0.603/M▼ 50.0%
new model
DeepSeek V4 Flash 0731

via scaleway

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B

via nano-gpt

What changed

  • Input price$0.13/M$0.08/M▼ 38.5%
  • Output price$0.4/M$0.33/M▼ 17.5%
  • Cache read$0.065/M$0.04/M▼ 38.5%
repriced
GLM-5

via hyper

What changed

  • Input price$0.83/M$0.91/M▲ 9.6%
  • Output price$2.56/M$2.93/M▲ 14.7%
  • cache write$0.415/M$0.455/M▲ 9.6%
repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$1.05/M$1.05/M▲ 0.2%
  • Output price$1.05/M$1.05/M▲ 0.2%
repriced
Kimi K2.5

via hyper

What changed

  • Input price$0.55/M$0.544/M▼ 1.1%
  • Output price$2.88/M$2.85/M▼ 1.0%
  • cache write$0.275/M$0.272/M▼ 1.1%
repriced
DeepSeek V4 Flash 0731

via edenai

What changed

  • Input price$0.467/M$0.468/M▲ 0.2%
  • Output price$0.934/M$0.936/M▲ 0.2%
capability
Ox Alpha Free (Unlimited)

via opencode

What changed

  • nameOx Alpha FreeOx Alpha Free (Unlimited)
new model
Ox Alpha Free (Unlimited)

via opencode-go

What changed

First observed in the model catalog

repriced
Gemini 3.6 Flash

via ofox

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
repriced
GPT-5.6 Sol

via github-copilot

What changed

  • Input price$5/M$2.50/M▼ 50.0%
  • Output price$30/M$15/M▼ 50.0%
  • Cache read$0.5/M$0.25/M▼ 50.0%
  • cache write$6.25/M$3.13/M▼ 50.0%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▲ 0.2%
  • Output price$0.701/M$0.702/M▲ 0.2%
repriced
DeepSeek V4 Flash

via openrouter

What changed

  • Input price$0.083/M$0.081/M▼ 1.9%
  • Output price$0.165/M$0.162/M▼ 1.9%
  • Cache read$0.017/M$0.016/M▼ 1.9%
context
DeepSeek V4 Flash 0731

via kilo

What changed

  • Output limit393K384K−2%
new model
DeepSeek V4 Flash Vision Exp

via opencode-go

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via edenai

What changed

First observed in the model catalog

new model
Qwen 3.8 27B Uncensored

via nano-gpt

What changed

First observed in the model catalog

new model
Qwen3.8 27B

via ofox

What changed

First observed in the model catalog

new model
DeepSeek V4 Pro 0423

via ofox

What changed

First observed in the model catalog

repriced
Gemma 4 26B A4B IT

via hyper

What changed

  • Input price$0.12/M$0.11/M▼ 8.3%
  • Output price$0.42/M$0.408/M▼ 2.9%
  • cache write$0.06/M$0.055/M▼ 8.3%
context
Nemotron 3 Super 120B A12B

via edenai

What changed

  • Context8K262K32.8×
repriced
DeepSeek V4 Flash

via edenai

What changed

  • Input price$0.44/M$0.22/M▼ 50.0%
  • Output price$1.32/M$0.66/M▼ 50.0%
  • Cache read$0.014/M$0.0070/M▼ 50.0%
repriced
Kimi K2.6

via openrouter

What changed

  • Input price$0.95/M$0.58/M▼ 39.0%
  • Output price$4/M$2.44/M▼ 39.0%
  • Cache read$0.16/M$0.098/M▼ 39.0%
new model
Ox Alpha

via venice

What changed

First observed in the model catalog

capability
DeepSeek V4 Flash Vision Exp

via kilo

What changed

  • nameDeepSeek: DeepSeek V4 Flash Vision ExpDeepSeek V4 Flash Vision Exp
  • familydeepseekdeepseek-flash
repriced
DeepSeek V4 Pro

via edenai

What changed

  • Input price$1.32/M$0.66/M▼ 50.0%
  • Output price$3.96/M$1.98/M▼ 50.0%
  • Cache read$0.044/M$0.022/M▼ 50.0%
repriced
GPT OSS 120B

via edenai

What changed

  • Input price$0.175/M$0.175/M▲ 0.2%
  • Output price$0.759/M$0.76/M▲ 0.2%
repriced
Gemma 4 31B

via nano-gpt

What changed

  • Input price$0.1/M$0.08/M▼ 20.0%
  • Output price$0.35/M$0.33/M▼ 5.7%
  • Cache read$0.05/M$0.04/M▼ 20.0%
new model
DeepSeek V4 Flash (RanoAI)

via llmgateway-providers

What changed

First observed in the model catalog

repriced
DeepSeek V4 Flash 0731

via openrouter

What changed

  • Output limit393K384K−2%
  • Input price$0.14/M$0.08/M▼ 42.9%
  • Output price$0.28/M$0.18/M▼ 35.7%
  • Cache read$0.028/M$0.016/M▼ 42.9%
repriced
Hy3

via kilo

What changed

  • Input price$0.14/M$0.132/M▼ 5.7%
  • Output price$0.58/M$0.528/M▼ 9.0%
  • Cache read$0.035/M$0.033/M▼ 5.7%
capability
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

  • familydeepseekdeepseek-flash
repriced
Gemini 3.7 Flash

via ofox

What changed

  • Input price$1.50/M$0.75/M▼ 50.0%
  • Output price$7.50/M$3.75/M▼ 50.0%
  • Cache read$0.15/M$0.075/M▼ 50.0%
  • cache write$0.083/M$0.042/M▼ 50.0%
new model
DeepSeek V4 Flash Vision Exp

via openrouter

What changed

First observed in the model catalog

new model
DeepSeek V4 Flash Vision Exp

via ofox

What changed

First observed in the model catalog

repriced
Llama-3.3-70B-Instruct

via edenai

What changed

  • Input price$0.759/M$0.76/M▲ 0.2%
  • Output price$0.759/M$0.76/M▲ 0.2%
new model
DeepSeek V4 Pro 0423

via merge-gateway

What changed

First observed in the model catalog

new model
Qwen 3.6 35B A3B Uncensored

via nano-gpt

What changed

First observed in the model catalog

repriced
DeepSeek V4 Pro 0813

via edenai

What changed

  • Input price$1.21/M$0.603/M▼ 50.0%
  • Output price$3.62/M$1.81/M▼ 50.0%
new model
Meituan: LongCat 2.0 (free)

via kilo

What changed

First observed in the model catalog