Claude 3.7 Sonnet
#112 of 274 scored.
AnthropicFeb 24, 2025Closed
Capabilities Index
Range 138.8–142.7
141.2
Context
Tokens
—
Input
Per 1M tokens
—
Output
Per 1M tokens
—
Benchmarks
How it scores.
- GPQA Diamond73.0%
81st of 186
- SWE-bench Verified61.0%
28th of 32
- ARC-AGI-20.9%
66th of 79
- Humanity's Last Exam3.4%
29th of 40
- Mock AIME57.7%
107th of 176
Everything else Epoch has for it
- MATH level 591.2%
- Fiction.LiveBench83.3%
- Lech Mazur Writing81.1%
- GeoBench68.0%
- Aider polyglot64.9%
- METR Time Horizons60.0%
- CadEval54.0%
- DeepResearch Bench43.6%
- OSWorld35.8%
- SimpleBench35.7%
- The Agent Company30.9%
- ARC-AGI28.6%
- Cybench20.0%
- VPCT8.5%
- FrontierMath-2025-02-28-Private7.3%
- GSO-Bench3.8%
On the index
Around it.
Sources