Anthropic · model
Claude Sonnet 5.5 benchmark scores
As of 2026-10-07, Claude Sonnet 5.5 (Anthropic) has measured scores on 33 benchmarks and estimated scores on 65 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 96.2% | estimated ± 2.1 pp, low confidence |
| AA-HLE | 55.0% | measured |
| AA-SciCode | 61.0% | measured |
| Artificial Analysis Intelligence Index | 56.0% | measured |
| CritPt | 31.4% | measured |
| GDPval-AA | 67.0% | measured |
| GPQA Diamond | 96.0% | estimated ± 1.0 pp, low confidence |
| HLE | 63.7% | estimated ± 5.7 pp, medium confidence |
| AA-LCR | 82.7% | measured |
| LiveCodeBench v6 | 99.0% | estimated ± 2.1 pp, medium confidence |
| AA-Omniscience Accuracy | 54.0% | measured |
| SciCode | 60.3% | estimated ± 3.2 pp, medium confidence |
| Terminal-Bench 2.1 | 89.9% | estimated ± 3.9 pp, medium confidence |
| BrowseComp | 87.0% | estimated ± 6.4 pp, medium confidence |
| HLE w/ tools | 64.5% | measured |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| τ²-bench results | 94.3% | estimated ± 7.1 pp, low confidence |
| AA Coding Index | 83.0% | estimated ± 1.0 pp, medium confidence |
| AA-MMMU-Pro | 83.4% | estimated ± 2.9 pp, high confidence |
| APEX-Agents | 51.5% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 48.1% | estimated ± 9.1 pp, low confidence |
| BioMysteryBench (human-difficult) | 44.7% | measured |
| DeepSearchQA | 90.5% | estimated ± 8.4 pp, low confidence |
| SWE-bench Verified | 88.3% | estimated ± 0.8 pp, medium confidence |
| AutomationBench | 40.4% | estimated ± 7.6 pp, low confidence |
| BFCL v4 | 78.5% | estimated ± 2.6 pp, medium confidence |
| CyberGym | 88.7% | estimated ± 2.4 pp, low confidence |
| JobBench | 65.2% | estimated ± 3.4 pp, low confidence |
| SWE-bench Pro | 81.3% | measured |
| WideResearch | 86.4% | estimated ± 3.4 pp, low confidence |
| AA Agentic Index | 60.3% | estimated ± 2.8 pp, medium confidence |
| MMLU-Pro | 86.0% | estimated ± 2.5 pp, low confidence |
| GPQA | 96.0% | estimated ± 1.0 pp, low confidence |
| Gert Labs | 85.6% | estimated ± 8.9 pp, low confidence |
| AA-AnalystAgent | 57.5% | measured |
| AA EnterpriseOps-Gym | 56.3% | estimated ± 1.8 pp, low confidence |
| AA Harvey LAB | 93.1% | measured |
| ARC-AGI-1 | 97.2% | estimated ± 3.7 pp, high confidence |
| ARC-AGI-2 | 84.8% | estimated ± 13.0 pp, low confidence |
| CursorBench 3.1 | 71.1% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 63.2% | estimated ± 1.6 pp, medium confidence |
| FrontierCode 1.1 Main | 46.2% | measured |
| FrontierSWE v2 | 61.9% | measured |
| OfficeQA Pro | 65.6% | measured |
| OSWorld-Verified | 90.7% | estimated ± 8.7 pp, low confidence |
| Vals GPQA Diamond | 98.2% | estimated ± 1.5 pp, low confidence |
| Vals LiveCodeBench | 87.0% | estimated ± 1.9 pp, high confidence |
| Vals MMLU-Pro | 96.7% | estimated ± 1.5 pp, low confidence |
| Vals SWE-bench | 95.4% | estimated ± 1.4 pp, medium confidence |
| Terminal-Bench 2.1 (Vals AI) | 98.6% | estimated ± 4.6 pp, low confidence |
| VulcanBench v3 | 87.0% | estimated ± 0.7 pp, low confidence |
| AA AutomationBench | 71.8% | measured |
| GDP.pdf | 25.8% | measured |
| AA ITBench | 43.4% | estimated ± 3.4 pp, medium confidence |
| AA Tau3 Banking | 47.0% | estimated ± 4.4 pp, medium confidence |
| AA Terminal-Bench 2.1 | 86.8% | estimated ± 2.6 pp, medium confidence |
| AA Terminal-Bench 4.0 | 63.6% | measured |
| CursorBench 4.0 | 55.5% | measured |
| CWE-bench v1 | 65.0% | estimated ± 6.0 pp, medium confidence |
| DeepSWE | 71.0% | measured |
| HLE w/o tools | 56.9% | measured |
| OSWorld 2.0 | 45.2% | estimated ± 5.5 pp, low confidence |
| PostTrainBench v1.1 | 44.4% | estimated ± 4.3 pp, medium confidence |
| ProgramBench | 79.7% | measured |
| SWE Multilingual | 90.3% | measured |
| Terminal-Bench 4.0 | 70.6% | measured |
| Terminal-Bench-Science 0.1 | 59.9% | measured |
| Toolathlon-Verified | 77.8% | measured |
| Vibe Code Bench | 75.0% | estimated ± 5.6 pp, low confidence |
| ExploitGym | 23.2% | estimated ± 5.0 pp, low confidence |
| C-Eval | 97.0% | estimated ± 1.0 pp, low confidence |
| MCP Atlas | 88.0% | estimated ± 2.7 pp, low confidence |
| MMLU-ProX | 90.4% | estimated ± 1.7 pp, low confidence |
| MMLU-Redux | 97.6% | estimated ± 1.1 pp, low confidence |
| NL2Repo | 71.6% | estimated ± 6.4 pp, low confidence |
| NOVA-63 | 63.2% | estimated ± 1.6 pp, low confidence |
| QwenClawBench | 100.0% | estimated ± 6.0 pp, low confidence |
| SuperGPQA | 87.1% | estimated ± 7.5 pp, low confidence |
| Toolathlon | 55.9% | estimated ± 6.3 pp, low confidence |
| HealthBench Hard | 35.1% | estimated ± 7.7 pp, low confidence |
| LiveCodeBench Pro | 98.0% | estimated ± 6.0 pp, low confidence |
| MedXpertQA (Text) | 72.9% | estimated ± 8.9 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 81.3% | estimated ± 4.7 pp, low confidence |
| ResearchClawBench | 20.5% | estimated ± 2.2 pp, low confidence |
| ARC-AGI-3 | 1.7% | estimated ± 3.4 pp, high confidence |
| BioMysteryBench (human-solvable) | 89.2% | measured |
| DRACO | 87.0% | measured |
| FrontierCode 1.1 Extended | 59.1% | measured |
| HealthBench (raw) | 69.4% | measured |
| HealthBench (length-adjusted) | 65.4% | measured |
| HealthBench Professional | 69.2% | measured |
| HealthBench Professional (raw) | 77.1% | measured |
| HLE-Verified | 50.6% | estimated ± 6.0 pp, low confidence |
| LABBench2 | 86.2% | estimated ± 2.4 pp, medium confidence |
| Agents' Last Exam | 50.2% | estimated ± 6.3 pp, medium confidence |
| OpenHarmony Bench | 58.8% | estimated ± 2.4 pp, medium confidence |
| MMLU | 95.4% | estimated ± 1.1 pp, low confidence |