GPT-5.4
OpenAI · Released 5 Mar 2026 · gpt-5.4
$2.50
Input price / 1M tokens
#45 of 61, cheapest first
$15.00
Output price / 1M tokens
#46 of 61, cheapest first
1.05M
Context window
#2 of 71
88
Output tokens / second
#34 of 61
75.1%
Terminal-Bench 2.0
#2 of 7
57.7%
SWE-bench Pro
#13 of 16
75%
OSWorld-Verified
#6 of 8
92.8%
GPQA Diamond
#7 of 27
Summary
GPT-5.4 is a proprietary model from OpenAI, released on 5 Mar 2026.
At $2.50 input and $15.00 output per million tokens, it is #45 of 61 on input price, cheapest first.
Its 1.05M-token context window ranks #2 of 71.
Measured output speed is 88 tokens per second, #34 of 61.
A frontier model for professional work and coding, and OpenAI's first general-purpose model with native computer use. It is now a lower-cost alternative to GPT-5.5.
Benchmarks · 6 reported
| Benchmark | Score | Bar | Rank | Source |
|---|---|---|---|---|
| Terminal-Bench 2.0xhigh effort | 75.1% | #2 of 7 | OpenAI GPT-5.5 launch post (comparison column)Self-reported | |
| SWE-bench Proxhigh effort | 57.7% | #13 of 16 | OpenAI GPT-5.5 launch post (comparison column)Self-reported | |
| OSWorld-Verifiedxhigh effort | 75% | #6 of 8 | OpenAI GPT-5.5 launch post (comparison column)Self-reported | |
| GPQA Diamondxhigh effort | 92.8% | #7 of 27 | OpenAI GPT-5.5 launch post (comparison column)Self-reported | |
| BrowseCompxhigh effort | 82.7% | #7 of 7 | OpenAI GPT-5.5 launch post (comparison column)Self-reported | |
| ARC-AGI-2xhigh effort | 73.3% | #4 of 6 | OpenAI GPT-5.5 launch post (comparison column)Self-reported |
Self-reported means the lab ran the test itself. A rank appears only where several labs report the same version of a benchmark.
Similar models
Models with the most benchmarks in common with GPT-5.4, and the closest scores.
Terminal-Bench 2.0
#2 of 7Terminal tasks, 2025 set
- GPT-5.475.1
- GPT-5.582.7
- Gemini 3.1 Pro68.5
- Kimi K2.666.7
SWE-bench Pro
#13 of 16Harder multi-file engineering tasks
- GPT-5.457.7
- GPT-5.558.6
- Gemini 3.1 Pro54.2
- Kimi K2.658.6
- Gemini 3.5 Flash53.9
- GPT-5.6 Sol64.6
OSWorld-Verified
#6 of 8Using a desktop computer
- GPT-5.475
- GPT-5.578.7
- Kimi K2.673.1
- Gemini 3.5 Flash78.4
GPQA Diamond
#7 of 27Graduate-level science questions
- GPT-5.492.8
- GPT-5.593.6
- Gemini 3.1 Pro94.3
- Kimi K2.690.5
- GPT-5.6 Sol94.6
Compare GPT-5.4 side by side
Pre-filled with the two closest models. Swap any of them.
Sources
- price, tiers, long-context rule: developers.openai.com/api/docs/pricing, checked 2 Oct 2026
- context, max output, knowledge cutoff (Aug 31 2025), snapshot, modalities: developers.openai.com/api/docs/models/gpt-5.4, checked 2 Oct 2026
- release date (changelog Mar 5 2026): developers.openai.com/api/docs/changelog, checked 2 Oct 2026
- benchmarks: openai.com/index/introducing-gpt-5-5/, checked 2 Oct 2026
- output speed: artificialanalysis.ai/models/gpt-5-4, checked 2 Oct 2026
Building on GPT-5.4?
Get the architecture right before the bill arrives.
We size caching, routing and fallbacks for your workload on a free call, and tell you where a cheaper model is good enough.