Skip to content
Model Pulse

Leaderboard

MCP Atlas

11 models scored · metric: success rate. Prices are the cheapest listed across all serving providers. Cost/run and Pts/$ are priced for the agentic token profile — tool loops re-send a growing transcript and think between calls.

A benchmark is priced for the workload it exercises, not a fixed input:output blend: an agent loop and a single hard question bill very differently on the same rate card. Profile: 12k in · 2k out · 3k reasoning · 70% cache hit · 32k context.

Scores are facts as publicly reported by labs (see report links below and on each model page) and aggregated via models.dev. MCP Atlas is maintained by its own project — Model Pulse is not affiliated with or endorsed by it.

Best value · top 5 by pts per dollar · agentic profile
  1. 1.Muse Glimmer 30B13,727
  2. 2.MiniMax-M312,178
  3. 3.GLM-5.210,535
  4. 4.Gemini 3.5 Flash10,278
  5. 5.GPT-5.59,169
11 models
#
🥇Muse Spark 1.1meta
88.1
$1.25$4.25$0.0293,0111.05M
🥈Gemini 3.5 Flashgoogle
83.6
$0.186$1.110.813¢10,2781.05M
🥉Gemini 3.1 Pro Previewgoogle
78.2
$1$6$0.0352,2491.05M
4GLM-5.2zhipuai
76.8
$0.3$1.050.729¢10,5351.05M
5Qwen3.7 Maxalibaba
76.4
$0.825$2.48$0.0184,1941.06M
6Kimi K2.7 Codemoonshotai
76
$0.275$1.100.930¢8,176271K
7Muse Glimmer 30Bmeta
75.5
$0.2$0.80.550¢13,727131K
8GPT-5.5openai
75.3
$0.188$1.130.821¢9,1691.05M
9MiniMax-M3minimax
74.2
$0.225$0.90.609¢12,1781.05M
10GPT-5.4 miniopenai
57.7
$0.375$4$0.0222,583400K
11GPT-5.4 nanoopenai
56.1
$0.18$1.100.662¢8,4701.05M
1–11 of 11 models

Score reports: ai.meta.com · deepmind.google · z.ai · qwen.ai · huggingface.co · huggingface.co — full source URLs are linked on each model page.

Cheapest scorer: GPT-5.4 nano at $0.18 input /M.