GPT-4.5
#134 of 274 scored.
OpenAIFeb 27, 2025Closed
Capabilities Index
Range 133.4–138.7
136.7
Context
Tokens
—
Input
Per 1M tokens
—
Output
Per 1M tokens
—
Benchmarks
How it scores.
- GPQA Diamond58.3%
103rd of 186
- ARC-AGI-20.8%
68th of 79
- Humanity's Last Exam0.7%
34th of 40
- Mock AIME37.7%
117th of 176
Everything else Epoch has for it
- MATH level 578.6%
- Lech Mazur Writing75.6%
- Fiction.LiveBench63.9%
- Aider polyglot44.9%
- WeirdML39.4%
- SimpleBench21.4%
- Cybench17.5%
- VPCT17.5%
- ARC-AGI10.3%
On the index
Around it.
Sources