Anthropic · model
Claude 4 Opus benchmark scores
As of 2026-10-29, Claude 4 Opus (Anthropic) has measured scores on 9 benchmarks and estimated scores on 53 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 70.1% | measured |
| AA-HLE | 6.2% | measured |
| AA-SciCode | 44.8% | estimated ± 3.9 pp, high confidence |
| AIME26 | 38.2% | estimated ± 4.6 pp, low confidence |
| CritPt | 0.2% | estimated ± 2.4 pp, high confidence |
| GPQA Diamond | 75.1% | estimated ± 1.2 pp, high confidence |
| HLE | 12.1% | estimated ± 6.9 pp, medium confidence |
| IFBench | 54.6% | estimated ± 6.6 pp, medium confidence |
| SciCode | 44.5% | estimated ± 3.3 pp, high confidence |
| IFEval | 89.8% | estimated ± 2.8 pp, high confidence |
| AA Coding Index | 44.8% | estimated ± 5.3 pp, medium confidence |
| AA-MMMU-Pro | 59.1% | estimated ± 3.0 pp, high confidence |
| SWE-bench Verified | 72.5% | measured |
| SWE-bench Pro | 50.0% | estimated ± 4.6 pp, high confidence |
| MMLU-Pro | 85.5% | estimated ± 4.1 pp, high confidence |
| AA-IFBench | 43.3% | measured |
| AIME 2024 | 49.4% | estimated ± 2.6 pp, high confidence |
| Artificial Analysis Intelligence Index | 16.6% | measured |
| GPQA | 74.9% | measured |
| MMLU | 91.8% | estimated ± 2.7 pp, high confidence |
| AA-Omniscience Accuracy | 26.8% | estimated ± 7.2 pp, medium confidence |
| MATH-500 | 91.5% | estimated ± 2.1 pp, high confidence |
| MMMLU | 87.4% | measured |
| MMMU | 73.7% | measured |
| AA MMLU-Pro | 85.4% | estimated ± 1.0 pp, low confidence |
| AIME 2025 | 33.9% | measured |
| ARC-AGI-1 | 31.0% | estimated ± 7.0 pp, low confidence |
| ARC-AGI-2 | 0.0% | estimated ± 11.6 pp, low confidence |
| FrontierCode 1.1 Main | 6.4% | estimated ± 2.7 pp, low confidence |
| Vals GPQA Diamond | 62.7% | estimated ± 4.8 pp, medium confidence |
| Vals LiveCodeBench | 72.2% | estimated ± 9.9 pp, low confidence |
| Vals MMLU-Pro | 46.9% | estimated ± 1.2 pp, medium confidence |
| Vals SWE-bench | 66.0% | estimated ± 6.8 pp, medium confidence |
| DeepSWE | 30.1% | estimated ± 12.7 pp, low confidence |
| HLE w/o tools | 31.1% | estimated ± 4.6 pp, high confidence |
| Vibe Code Bench | 1.0% | estimated ± 8.1 pp, medium confidence |
| ARC-AGI-3 | 1.4% | estimated ± 7.7 pp, low confidence |
| AI2D_TEST | 89.2% | estimated ± 2.0 pp, high confidence |
| C-Eval | 90.5% | estimated ± 1.0 pp, medium confidence |
| CountBench | 94.5% | estimated ± 3.2 pp, high confidence |
| DynaMath | 82.5% | estimated ± 5.8 pp, low confidence |
| ERQA | 47.1% | estimated ± 5.2 pp, low confidence |
| MathVision | 72.5% | estimated ± 3.6 pp, high confidence |
| MLVU (M-Avg) | 83.0% | estimated ± 2.4 pp, high confidence |
| MMLU-Redux | 94.1% | estimated ± 0.5 pp, medium confidence |
| MMMU-Pro | 62.1% | estimated ± 2.3 pp, high confidence |
| NL2Repo | 30.6% | estimated ± 6.2 pp, medium confidence |
| RealWorldQA | 80.2% | estimated ± 3.6 pp, high confidence |
| ScreenSpot Pro | 53.0% | estimated ± 7.3 pp, low confidence |
| SimpleVQA | 55.1% | estimated ± 7.1 pp, medium confidence |
| SuperGPQA | 55.2% | estimated ± 8.2 pp, low confidence |
| Video-MME (w/o subtitle) | 74.8% | estimated ± 1.3 pp, high confidence |
| Video-MME (with subtitle) | 80.3% | estimated ± 3.2 pp, high confidence |
| VideoMMMU | 68.3% | estimated ± 3.8 pp, high confidence |
| We-Math | 76.7% | estimated ± 7.2 pp, low confidence |
| React Native Evals | 72.4% | estimated ± 2.3 pp, medium confidence |
| SWE-Rebench | 57.0% | estimated ± 3.6 pp, medium confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 3.3 pp, medium confidence |
| HealthBench Professional (raw) | 53.8% | estimated ± 8.6 pp, low confidence |
| SimpleQA | 23.6% | estimated ± 7.8 pp, medium confidence |
| LiveCodeBench | 74.6% | estimated ± 14.6 pp, low confidence |
| MMVU | 68.9% | estimated ± 3.8 pp, medium confidence |