RSS

OpenAI

GPT-5.4

Benchmarks, price, context and versions

Also searched as

GPT 5.4, gpt-5.4-mini, gpt-5.4 pricing, gpt-5.4-nano

Direct answer: GPT-5.4 is a model from OpenAI. 93.3% on GPQA Diamond, per Epoch AI. On OpenRouter, US$ 2.50 per 1 million input tokens and US$ 15.00 per 1 million output tokens.

✓ Good for

  • technical reasoning on science topics (#20 of 198 on GPQA Diamond)
  • calculation and problems with an exact answer (#33 of 186 on OTIS Mock AIME)
  • the ceiling of reasoning on research-level math (#21 of 86 on FrontierMath (tiers 1 to 3))
  • fixing bugs in real code (#8 of 32 on SWE-bench Verified)
  • very long documents or code (a window of 1 million tokens or more, according to OpenRouter)

✕ Not the best for

  • when the input budget sits at the top of the table (input price in the highest 25% on OpenRouter)

This block follows a rule of this section, using source data: if the model is in the top 25% of the metric, the item goes under "Good for"; if it is in the bottom half, it goes under "Not the best for". The middle band appears in neither. Price and context use OpenRouter; scores use Epoch AI or OpenRouter Benchmarks, in the latest version we checked.

Benchmarks

GPQA Diamond

?what it is
93.3%

highest score on this metric 96.4%

20º / 198 · Epoch AI, latest version we checked

OTIS Mock AIME

?
97.8%

highest score on this metric 100%

33º / 186 · Epoch AI, latest version we checked

FrontierMath (tiers 1 to 3)

?
78.6%

highest score on this metric 95.1%

21º / 86 · Epoch AI, latest version we checked

SimpleQA Verified

?
45.1%

highest score on this metric 94%

41º / 87 · Epoch AI, latest version we checked

SWE-bench Verified

?
76.9%

highest score on this metric 83.5%

8º / 32 · Epoch AI, latest version we checked

GPQA Diamond (OpenRouter)

?
90.3%

highest score on this metric 96.5%

36º / 135 · OpenRouter Benchmarks, latest version we checked

τ-bench Verified (airline)

?
75%

highest score on this metric 82%

52º / 131 · OpenRouter Benchmarks, latest version we checked

Key specs

Origin
the US
Family
GPT provisional data
Weights?
closed
License
no data
Input US$/1M?
US$ 2.50
Output US$/1M
US$ 15.00
Context?
1,050,000 tokens
Parameters
no data
Input
file, image, text
OpenRouter usage?
no data
Downloads (30 days)?
no data
Use cases provisional data
Text Agents Code
Dates in sources
  • 2026, published, per Epoch AI
  • 2026, listed on OpenRouter

Version history

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch.

FAQ

How much does GPT-5.4 cost?

On OpenRouter, US$ 2.50 per 1 million input tokens and US$ 15.00 per 1 million output tokens.

What is the context window of GPT-5.4?

1,050,000 tokens, per OpenRouter.

What is GPT-5.4?

GPT-5.4 is a model from OpenAI, with closed weights, used for text, agents, code.

Is GPT-5.4 open source?

No. It is available through API only.

What is GPT-5.4's GPQA Diamond score?

93.3% on GPQA Diamond, per Epoch AI.

Is GPT-5.4 good for coding?

It scores 76.9% on SWE-bench Verified, per Epoch AI.

Who made GPT-5.4?

It was made by OpenAI, a lab from the US.

provisional data Questions ordered by what shows up in public Google search suggestions (no volume). Answers use source data only.

Models from OpenAI

See lab

Sources

  • Epoch AI, CC BY 4.0, latest version we checked.
  • OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.
  • OpenRouter Benchmarks, CC BY 4.0, latest version we checked.

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.