Claude Opus 4.7
#25 of 274 scored.
AnthropicApr 16, 2026Closed
Capabilities Index
Range 154.4–158.4
156.3
Context
Takes text, image, file
1M
Input
Per 1M tokens
$5.00
Output
Per 1M tokens
$25
Benchmarks
How it scores.
- GPQA Diamond86.9%
39th of 186
- SWE-bench Verified83.5%
1st of 32
- FrontierMath70.2%
23rd of 81
- ARC-AGI-275.8%
15th of 79
- Humanity's Last Exam33.0%
9th of 40
- SimpleQA Verified51.7%
23rd of 77
- Mock AIME97.8%
23rd of 176
- Terminal-Bench80.2%
3rd of 35
Everything else Epoch has for it
- ARC-AGI93.5%
- DTBench91.1%
- WeirdML76.4%
- LMCA61.4%
- SimpleBench54.0%
- ProofBench54.0%
- APEX-Agents49.2%
- GSO-Bench44.1%
- FrontierCode38.5%
- FrontierMath-Tier-4-v2-Private31.7%
- MirrorCode31.1%
- PostTrainBench28.6%
- ExploitBench26.5%
- Chess Puzzles26.4%
- Mystery Game Puzzles20.7%
- EBR-bench19.1%
- OSWorld 2.018.2%
- Furniture Assembly4.8%
On the index
Around it.
Sources