o3
#69 of 274 scored.
OpenAIApr 16, 2025Closed
Capabilities Index
Range 145.0–148.7
146.9
Context
Takes image, text, file
200K
Input
Per 1M tokens
$2.00
Output
Per 1M tokens
$8.00
Benchmarks
How it scores.
- GPQA Diamond75.8%
78th of 186
- SWE-bench Verified62.3%
27th of 32
- FrontierMath33.3%
59th of 81
- ARC-AGI-26.5%
47th of 79
- Humanity's Last Exam16.3%
18th of 40
- SimpleQA Verified49.4%
26th of 77
- Mock AIME84.4%
71st of 176
Everything else Epoch has for it
- MATH level 597.8%
- Fiction.LiveBench88.9%
- Lech Mazur Writing83.9%
- Aider polyglot81.3%
- DTBench74.7%
- CadEval74.0%
- GeoBench74.0%
- METR Time Horizons65.4%
- ARC-AGI60.8%
- WeirdML52.4%
- LMCA46.7%
- DeepResearch Bench45.2%
- SimpleBench43.7%
- Chess Puzzles34.8%
- GDPval30.8%
- VPCT28.0%
- OSWorld23.0%
- Mystery Game Puzzles21.8%
- CL-bench17.8%
- GSO-Bench8.8%
- FrontierMath-Tier-4-2025-07-01-Private3.5%
On the index
Around it.
Sources