in the loop

GPT-5 Pro
#51 of 274 scored.

OpenAIOct 7, 2025Closed

Capabilities Index

Range 148.6–152.8

150.3

Context

Takes image, text, file

400K

Input

Per 1M tokens

$15

Output

Per 1M tokens

$120

Benchmarks

How it scores.

  • FrontierMath55.8%

    40th of 81

  • ARC-AGI-218.3%

    39th of 79

  • Humanity's Last Exam28.2%

    11th of 40

Everything else Epoch has for it

  • ARC-AGI70.2%
  • WeirdML60.4%
  • SimpleBench53.9%
  • FrontierMath-Tier-4-v2-Private19.5%

On the index

Around it.

  1. #48Gemini 3 FlashGoogle151.8
  2. #49GLM-5.2Z.ai · $0.06 in151.8
  3. #50Kimi K2.6Moonshot · $0.47 in151.1
  4. #51GPT-5 ProOpenAI · $15 in150.3
  5. #52Inkling-SmallThinking Machines · $0.45 in150.2
  6. #53Claude Opus 4.5Anthropic · $5.00 in150.1
  7. #54GPT-5OpenAI · $1.25 in150.0

Sources