OpenAI
OpenAI
AI model · United States ·Mar 2026

GPT-5.4 Thinking

Proprietary
1.1M
Context
4
Benchmarks

About GPT-5.4 Thinking

GPT-5.4 Thinking is an AI model developed by OpenAI, released in 2026, made available as a proprietary (API-only) model. GPT-5.4 Thinking is the reasoning-optimized variant of GPT-5.4, released March 5, 2026. The System Card shows it is the .

On this page you'll find GPT-5.4 Thinking's full specifications, including a context window of 1.1M tokens, 4 benchmark results. Review provider pricing and benchmark scores below, or compare GPT-5.4 Thinking head-to-head with other AI models.

→ More models by OpenAI → AI benchmarks leaderboard

Links

Announcement ↗ Paper ↗

Community

Would you run GPT-5.4 Thinking again?
No votes yet — be the first.
Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

GPT-5.4 Thinking benchmark scores▶
Compare on the benchmark table →
Benchmark Category Score Variant
SWE-bench Verified code 74.9% thinking
AIME 2025 math 88.5% thinking
GPQA Diamond reasoning 85.7% thinking
Humanity's Last Exam reasoning 24.8% thinking

More models like GPT-5.4 Thinking

OpenAI
GPT-5.5-Cyber
OpenAI
 OpenAI
GPT-Realtime-Whisper
OpenAI
 OpenAI
GPT-Realtime-2
OpenAI
 OpenAI
GPT-Realtime-Translate
OpenAI
 OpenAI
GPT-5.5 Instant
OpenAI
 OpenAI
GPT-5.5
OpenAI

Frequently asked questions

What is GPT-5.4 Thinking? ▶

GPT-5.4 Thinking is the reasoning-optimized variant of GPT-5.4, released March 5, 2026. The System Card shows it is the It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created GPT-5.4 Thinking? ▶

GPT-5.4 Thinking was developed by OpenAI and released in 2026.

Is GPT-5.4 Thinking open source or proprietary? ▶

GPT-5.4 Thinking is a proprietary model. It is accessed through an API rather than by downloading the weights.

What is GPT-5.4 Thinking's context window? ▶

GPT-5.4 Thinking supports a context window of 1.1M tokens.

How much does GPT-5.4 Thinking cost? ▶

Pricing for GPT-5.4 Thinking depends on the provider. See the providers table on this page for the latest API rates.

How does GPT-5.4 Thinking perform on benchmarks? ▶

GPT-5.4 Thinking is benchmarked across 4 evaluations on GenAIList, including SWE-bench Verified (74.9%). See the full benchmark table below and compare it with other models.

GPT-5.4 Is HERE – Hands-On With OpenAI's Newest Model!

Know a great video about GPT-5.4 Thinking?