Anthropic · model
Claude Haiku 5.5 benchmark scores
As of 2026-10-07, Claude Haiku 5.5 (Anthropic) has measured scores on 13 benchmarks and estimated scores on 65 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 91.4% | estimated ± 2.7 pp, high confidence |
| AA-HLE | 44.4% | measured |
| AA-SciCode | 55.0% | measured |
| Artificial Analysis Intelligence Index | 43.4% | measured |
| CritPt | 18.9% | measured |
| GDPval-AA | 41.8% | estimated ± 6.9 pp, medium confidence |
| GPQA Diamond | 93.9% | estimated ± 1.9 pp, high confidence |
| HLE | 54.8% | estimated ± 5.7 pp, medium confidence |
| AA-LCR | 82.7% | measured |
| LiveCodeBench v6 | 95.9% | estimated ± 3.3 pp, medium confidence |
| AA-Omniscience Accuracy | 46.1% | estimated ± 6.9 pp, medium confidence |
| SciCode | 54.5% | estimated ± 3.2 pp, high confidence |
| Terminal-Bench 2.1 | 88.4% | estimated ± 3.9 pp, high confidence |
| BrowseComp | 84.6% | estimated ± 7.6 pp, medium confidence |
| HLE w/ tools | 57.4% | measured |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| AA Coding Index | 75.2% | estimated ± 3.3 pp, high confidence |
| BioMysteryBench (human-difficult) | 47.8% | estimated ± 6.8 pp, low confidence |
| SWE-bench Verified | 88.4% | estimated ± 0.8 pp, medium confidence |
| AutomationBench | 44.0% | estimated ± 7.6 pp, low confidence |
| CyberGym | 80.1% | estimated ± 2.4 pp, medium confidence |
| JobBench | 0.7% | estimated ± 9.1 pp, low confidence |
| SWE-bench Pro | 67.0% | estimated ± 6.8 pp, medium confidence |
| WideResearch | 82.8% | estimated ± 3.4 pp, low confidence |
| AA Agentic Index | 37.1% | estimated ± 6.7 pp, medium confidence |
| MMLU-Pro | 85.8% | estimated ± 2.5 pp, low confidence |
| GPQA | 93.6% | estimated ± 2.1 pp, high confidence |
| AA-AnalystAgent | 51.5% | estimated ± 2.5 pp, medium confidence |
| AA EnterpriseOps-Gym | 45.8% | estimated ± 4.4 pp, high confidence |
| AA Harvey LAB | 89.9% | measured |
| ARC-AGI-1 | 95.5% | estimated ± 3.7 pp, high confidence |
| ARC-AGI-2 | 78.0% | estimated ± 13.0 pp, low confidence |
| CursorBench 3.1 | 56.3% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 63.4% | estimated ± 1.6 pp, medium confidence |
| FrontierCode 1.1 Main | 46.4% | measured |
| FrontierSWE v2 | 27.1% | estimated ± 14.6 pp, low confidence |
| Vals GPQA Diamond | 91.5% | estimated ± 3.0 pp, high confidence |
| Vals LiveCodeBench | 87.1% | estimated ± 1.9 pp, high confidence |
| Vals MMLU-Pro | 89.1% | estimated ± 1.9 pp, high confidence |
| Vals SWE-bench | 86.6% | estimated ± 2.8 pp, high confidence |
| Terminal-Bench 2.1 (Vals AI) | 77.4% | estimated ± 4.6 pp, medium confidence |
| AA AutomationBench | 35.4% | measured |
| GDP.pdf | 20.8% | measured |
| AA ITBench | 49.6% | estimated ± 3.4 pp, medium confidence |
| AA Tau3 Banking | 34.8% | estimated ± 7.1 pp, medium confidence |
| AA Terminal-Bench 2.1 | 86.6% | estimated ± 2.6 pp, high confidence |
| AA Terminal-Bench 4.0 | 32.8% | measured |
| CursorBench 4.0 | 39.0% | estimated ± 3.5 pp, high confidence |
| CWE-bench v1 | 39.6% | estimated ± 6.0 pp, medium confidence |
| DeepSWE | 69.3% | estimated ± 3.4 pp, medium confidence |
| HLE w/o tools | 45.9% | measured |
| OSWorld 2.0 | 36.8% | estimated ± 13.3 pp, low confidence |
| PostTrainBench v1.1 | 32.6% | estimated ± 5.9 pp, low confidence |
| Terminal-Bench 4.0 | 39.2% | measured |
| Terminal-Bench-Science 0.1 | 13.4% | estimated ± 4.1 pp, low confidence |
| Toolathlon-Verified | 76.6% | estimated ± 2.2 pp, medium confidence |
| Vibe Code Bench | 75.2% | estimated ± 5.6 pp, low confidence |
| ExploitGym | 16.4% | estimated ± 5.0 pp, low confidence |
| C-Eval | 96.0% | estimated ± 1.0 pp, low confidence |
| MCP Atlas | 80.2% | estimated ± 2.7 pp, medium confidence |
| MMLU-Redux | 95.2% | estimated ± 2.1 pp, low confidence |
| NL2Repo | 57.8% | estimated ± 6.5 pp, medium confidence |
| SuperGPQA | 91.4% | estimated ± 10.4 pp, low confidence |
| HealthBench Hard | 34.2% | estimated ± 7.7 pp, low confidence |
| MedXpertQA (Text) | 63.2% | estimated ± 8.9 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 78.8% | estimated ± 4.7 pp, medium confidence |
| ARC-AGI-3 | 1.6% | estimated ± 10.7 pp, low confidence |
| BioMysteryBench (human-solvable) | 88.9% | estimated ± 0.6 pp, medium confidence |
| HealthBench (raw) | 53.0% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 3.7 pp, medium confidence |
| HealthBench Professional | 54.5% | estimated ± 4.8 pp, medium confidence |
| HealthBench Professional (raw) | 63.1% | estimated ± 5.0 pp, medium confidence |
| HLE-Verified | 54.0% | estimated ± 6.8 pp, low confidence |
| LABBench2 | 81.7% | estimated ± 2.4 pp, medium confidence |
| Agents' Last Exam | 42.6% | estimated ± 6.3 pp, medium confidence |
| OpenHarmony Bench | 58.8% | estimated ± 2.4 pp, high confidence |
| MMLU | 100.0% | estimated ± 1.9 pp, low confidence |