Model

GPT-5.4

Name: GPT-5.4
Brand: openai

Agent-ready GPT for coding and computer-use workflows at a lower cost

43 providers · 47 listingsreleased Mar 5, 2026knowledge cutoff 2025-08-31closed weightsbeta

$0.75

lowest input /M · Xpersona

lowest output /M · Xpersona

1.05M

context window

131K

max output

Compare providers (sorted by listed input price)

40 of 43 providers publish a price for GPT-5.4. The cheapest full-context option is UnoRouter at $1.8/$10.8 per M — 23% below OpenAI's own price. Xpersona lists lower but caps context at 1.05M.

			Cache read /M	Cache write /M		Capabilities
Xpersona gpt-5.4	$0.75	$6	$0.075	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
UnoRoutervariant gpt-5.4	$1.80	$10.80	—	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Poe openai/gpt-5.4	$2.20	$14	$0.22	—	1.05M	reasoningtoolsno structuredvision	Feb 26, 2026
NanoGPT openai/gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Vivgrid gpt-5.4	$2.50	$15	$0.25	—	400K	reasoningtoolsstructuredvision	Mar 5, 2026
FreeModel gpt-5.4	$2.50	$15	$0.25	$2.50	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Impossibl openai/gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
LLM Gatewayvariant azure/gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
LLM Gatewayvariant openai/gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
AIHubMix gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
OrcaRouter openai/gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
SAP AI Core gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Azure Cognitive Services gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Pioneer gpt-5.4	$2.50	$15	$0.25	$2.50	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Requestyvariant gpt-5.4@eu	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
DaoXE gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
CrossModel openai/gpt-5.4	$2.50	$15	$0.25	$2.50	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Vercel AI Gatewayvariant openai/gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
DevPass (LLM Gateway) gpt-5.4	$2.50	$15	$0.25	—	1.05M	reasoningtoolsstructuredvision	Mar 5, 2026
Cloudflare AI Gateway openai/gpt-5.4	$2.50	$15	$0.25	—	1M	reasoningtoolsstructuredvision	Mar 5, 2026

Best price history

Daily lowest published price across live providers. A lower line is better for the buyer.

Best input /M: $0.75unchanged
Best output /M: $6unchanged
Provider coverage: 47; 46 at the start · 24 snapshots

Published best prices are unchanged.

24 daily snapshots collected from Aug 21, 2026 through Sep 14, 2026. A price chart will appear after the first real move.

Price badge

![GPT-5.4 price](https://www.vaanalytics.in/badge/openai/gpt-5.4.svg)

https://www.vaanalytics.in/badge/openai/gpt-5.4.svg

Live SVG, regenerated on every hourly sync — paste it into a README to always show the current cheapest listed price. Updates when prices change; no build step needed on your side.

Recent changes

Changelog →

contextvia nano-gpt · Sep 7, 2026
- context: 922K → 1.05M tokens
- input: 922K → 1.05M tokens
contextvia cortecs · Sep 2, 2026
- output: 1.05M → 128K tokens
repricedvia requesty · Aug 28, 2026
- input: $2.25 → $2.50
- output: $13.50 → $15
- cache_read: $0.225 → $0.25
- tiers: [object Object] → [object Object]
repricedvia orcarouter · Aug 28, 2026
- input: $5 → $2.50
- output: $22.50 → $15
repricedvia requesty · Aug 28, 2026
- input: $2.48 → $2.75
- output: $14.85 → $16.50
- cache_read: $0.247 → $0.275
capabilityvia azure-cognitive-services · Aug 27, 2026
- temperature: false → true

Benchmarks

SWE-Bench Pro ↗(resolve rate)report ↗Independent59.1
SWE-Atlas Codebase QnA(score)report ↗Independent40.8
SWE-Atlas Codebase QnA(score)report ↗Independent36.3
SWE-Atlas Refactoring(score)report ↗Independent44.29
SWE-Atlas Test Writing(score)report ↗Independent44.36
SWE-Atlas Test Writing(score)report ↗Independent40
Artificial Analysis Coding Agent Index ↗(average pass@1)report ↗Independent53.6
SWE-Atlas Codebase QnA(pass@1)report ↗Independent72.4
SWE-Bench Pro ↗(pass@1)report ↗Independent18.4
Terminal-Bench ↗(pass@1)report ↗Independent69.8
Artificial Analysis Coding Agent Index ↗(average pass@1)report ↗Independent52.2
SWE-Atlas Codebase QnA(pass@1)report ↗Independent72.9
SWE-Bench Pro ↗(pass@1)report ↗Independent18.9
Terminal-Bench ↗(pass@1)report ↗Independent64.7
Terminal-Bench ↗(success rate)report ↗Unclassified source75.1
GPQA Diamond ↗(accuracy)report ↗Unclassified source92.8
Humanity's Last Exam ↗(accuracy)report ↗Unclassified source39.8
Humanity's Last Exam ↗(accuracy)report ↗Unclassified source52.1
OSWorld-Verified ↗(success rate)report ↗Unclassified source75
BrowseComp ↗(accuracy)report ↗Unclassified source82.7
GDPval(wins or ties)report ↗Unclassified source83
ARC-AGI-2 ↗(accuracy)report ↗Unclassified source73.3
FrontierMath ↗(accuracy)report ↗Unclassified source47.6
FrontierMath ↗(accuracy)report ↗Unclassified source27.1
MMMU Pro ↗(accuracy)report ↗Unclassified source81.2

Scores a lab published about its own model are claims, not measurements. Rankings elsewhere on this site use independently measured scores only.

Prices are per million tokens (USD). “—” means the provider does not publicly list a price for this model. Data from models.dev; verify with the provider before purchasing.