Claude Opus 4.8
#14 of 274 scored.
AnthropicMay 28, 2026Closed
Capabilities Index
Range 156.0–160.8
158.2
Context
Takes text, image, file
1M
Input
Per 1M tokens
$5.00
Output
Per 1M tokens
$25
Benchmarks
How it scores.
- GPQA Diamond88.0%
28th of 186
- FrontierMath80.0%
15th of 81
- ARC-AGI-272.1%
17th of 79
- SimpleQA Verified53.0%
20th of 77
- Mock AIME98.3%
20th of 176
Everything else Epoch has for it
- ARC-AGI92.5%
- DTBench91.5%
- Surface Evolver Bench87.5%
- WeirdML82.9%
- ProofBench69.0%
- LMCA67.7%
- DeepSWE59.0%
- SimpleBench57.8%
- FrontierMath-Tier-4-v2-Private56.1%
- DeepResearch Bench50.2%
- APEX-Agents48.9%
- GSO-Bench47.1%
- FrontierCode46.5%
- PostTrainBench33.8%
- Chess Puzzles30.6%
- Mystery Game Puzzles29.5%
- EBR-bench28.6%
- OSWorld 2.020.6%
- Furniture Assembly17.9%
- Remote Labor Index8.3%
On the index
Around it.
Sources