Anthropic · model
Claude Opus 4.5 benchmark scores
As of 2026-11-22, Claude Opus 4.5 (Anthropic) has measured scores on 80 benchmarks and estimated scores on 115 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 81.0% | measured |
| AA-HLE | 13.2% | measured |
| AA-SciCode | 47.6% | estimated ± 3.1 pp, high confidence |
| AIME26 | 95.1% | measured |
| Artificial Analysis Intelligence Index | 23.7% | measured |
| CritPt | 0.3% | measured |
| GDPval-AA | 27.4% | estimated ± 5.0 pp, high confidence |
| GPQA Diamond | 87.0% | estimated ± 0.1 pp, medium confidence |
| HLE | 30.8% | measured |
| IFBench | 58.0% | measured |
| AA-LCR | 70.7% | measured |
| LiveCodeBench v6 | 84.8% | measured |
| AA-Omniscience Accuracy | 40.9% | measured |
| SciCode | 48.0% | estimated ± 3.6 pp, high confidence |
| BrowseComp | 73.4% | estimated ± 4.3 pp, high confidence |
| HLE w/ tools | 50.0% | estimated ± 4.9 pp, medium confidence |
| IFEval | 90.9% | measured |
| LongBench v2 | 64.4% | measured |
| VITA-Bench | 23.3% | measured |
| τ²-bench results | 86.3% | measured |
| AA Coding Index | 61.7% | estimated ± 4.3 pp, high confidence |
| AA-MMMU-Pro | 71.2% | measured |
| APEX-Agents | 0.0% | estimated ± 12.6 pp, low confidence |
| APEX-Agents-AA | 30.9% | estimated ± 3.6 pp, medium confidence |
| BioMysteryBench (human-difficult) | 6.5% | estimated ± 6.8 pp, low confidence |
| DeepSearchQA | 81.1% | estimated ± 7.3 pp, low confidence |
| SWE-bench Verified | 80.9% | measured |
| AutomationBench | 48.6% | estimated ± 9.1 pp, low confidence |
| BFCL v4 | 67.5% | estimated ± 4.7 pp, low confidence |
| CyberGym | 50.6% | measured |
| JobBench | 32.3% | measured |
| SkillsBench | 51.6% | estimated ± 4.9 pp, medium confidence |
| SWE-bench Pro | 57.1% | measured |
| τ³-bench results | 70.2% | measured |
| WideResearch | 76.4% | measured |
| AA Agentic Index | 18.8% | estimated ± 5.8 pp, low confidence |
| MMLU-Pro | 89.5% | measured |
| AA-IFBench | 43.0% | measured |
| BBH | 100.0% | estimated ± 2.8 pp, medium confidence |
| FailureSensorIQ | 66.7% | measured |
| GPQA | 87.0% | measured |
| MMLU | 90.5% | estimated ± 0.9 pp, medium confidence |
| Vals GPQA Diamond | 79.5% | measured |
| Vals LiveCodeBench | 75.0% | measured |
| Vals MMLU-Pro | 85.6% | measured |
| MATH-500 | 97.7% | estimated ± 3.1 pp, medium confidence |
| MMMLU | 85.4% | estimated ± 0.7 pp, medium confidence |
| MMMU | 80.7% | measured |
| Chess Puzzles | 14.0% | measured |
| AIME 2025 | 96.4% | estimated ± 1.8 pp, medium confidence |
| AA MMLU-Pro | 87.5% | estimated ± 0.7 pp, medium confidence |
| Gert Labs | 46.0% | measured |
| AA EnterpriseOps-Gym | 44.5% | estimated ± 1.2 pp, low confidence |
| AA Tau3 Banking | 39.6% | estimated ± 2.0 pp, low confidence |
| ApprenticeBench | 4.1% | estimated ± 4.0 pp, medium confidence |
| ARC-AGI-1 | 66.5% | estimated ± 6.5 pp, medium confidence |
| ARC-AGI-2 | 34.1% | estimated ± 10.9 pp, low confidence |
| BabaIsBench | 91.8% | estimated ± 1.2 pp, high confidence |
| CursorBench 3.1 | 52.7% | estimated ± 5.6 pp, low confidence |
| CursorBench 3.2 | 53.9% | estimated ± 3.6 pp, high confidence |
| EBR-bench | 14.3% | measured |
| FrontierCode 1.1 Main | 29.1% | estimated ± 4.2 pp, medium confidence |
| FrontierMath v2 (Tier 4) | 4.9% | measured |
| FrontierMath v2 (Tiers 1-3) | 34.4% | measured |
| FrontierSWE v2 | 10.6% | estimated ± 8.3 pp, medium confidence |
| Furniture Assembly | 28.3% | measured |
| MaxxBench | 42.0% | estimated ± 4.3 pp, low confidence |
| MazeBench | 0.4% | estimated ± 2.7 pp, high confidence |
| Mystery Game Puzzles | 22.0% | measured |
| NYT Connections | 49.4% | measured |
| OfficeQA Pro | 50.3% | estimated ± 6.8 pp, low confidence |
| OSWorld-Verified | 66.3% | measured |
| ParseBench-1 | 39.9% | estimated ± 5.2 pp, low confidence |
| Terminal-Bench Hard | 44.4% | estimated ± 4.0 pp, high confidence |
| Vals SWE-bench | 74.6% | estimated ± 2.8 pp, medium confidence |
| Terminal-Bench 2.1 (Vals AI) | 52.3% | estimated ± 6.4 pp, medium confidence |
| VulcanBench v3 | 79.1% | estimated ± 5.6 pp, low confidence |
| CWE-Bench-AA | 50.1% | estimated ± 8.1 pp, medium confidence |
| AA Cyber Index | 38.9% | estimated ± 12.4 pp, low confidence |
| DeepsecBench-AA | 15.5% | estimated ± 8.2 pp, medium confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 53.0% | estimated ± 4.3 pp, low confidence |
| HealthBench Professional (raw, Artificial Analysis) | 58.4% | estimated ± 1.1 pp, low confidence |
| AA ITBench | 27.1% | estimated ± 4.4 pp, medium confidence |
| Bug Hunt Bench | 12.1% | estimated ± 7.2 pp, low confidence |
| Chartography (tools) | 77.8% | estimated ± 4.1 pp, low confidence |
| CursorBench 4.0 | 33.4% | estimated ± 3.3 pp, medium confidence |
| CWE-bench v1 | 48.3% | estimated ± 5.7 pp, medium confidence |
| DeepSWE | 40.8% | estimated ± 8.7 pp, medium confidence |
| FrontierCode 2.0 | 20.3% | estimated ± 6.7 pp, low confidence |
| FrontierPhysics | 0.2% | estimated ± 4.4 pp, high confidence |
| HLE w/o tools | 32.3% | estimated ± 4.9 pp, high confidence |
| OSWorld 2.0 | 0.1% | estimated ± 3.6 pp, medium confidence |
| OSWorld 2.1 | 54.9% | estimated ± 3.6 pp, medium confidence |
| OSWorld 2.1 Offline | 55.4% | estimated ± 10.1 pp, low confidence |
| PostTrainBench v1.1 | 12.9% | estimated ± 1.3 pp, low confidence |
| SpireClimb | 5.8% | estimated ± 7.2 pp, low confidence |
| StealthMark | 45.3% | estimated ± 6.5 pp, medium confidence |
| SWE Multilingual | 77.5% | measured |
| Toolathlon-Verified | 71.3% | estimated ± 3.0 pp, high confidence |
| VulcanBench Frontier v4 | 78.0% | estimated ± 8.3 pp, low confidence |
| AA AutomationBench | 52.0% | estimated ± 9.9 pp, low confidence |
| GDP.pdf | 16.4% | estimated ± 4.1 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 8.3 pp, low confidence |
| ARC-AGI-3 Provider Adapter | 2.7% | estimated ± 7.3 pp, low confidence |
| CWE-bench | 44.4% | estimated ± 6.2 pp, medium confidence |
| MKA | 34.6% | estimated ± 4.6 pp, high confidence |
| SkillsBench official leaderboard (BenchFlow) | 49.0% | measured |
| BinaryAudit | 36.8% | measured |
| CompileBench | 88.9% | estimated ± 10.0 pp, low confidence |
| OTelBench | 29.0% | measured |
| Vibe Code Bench | 55.5% | estimated ± 8.4 pp, medium confidence |
| FACTS Multimodal | 29.0% | estimated ± 4.0 pp, medium confidence |
| Hermes Index | 30.1% | estimated ± 5.7 pp, low confidence |
| ParseBench-2 | 59.0% | estimated ± 5.0 pp, medium confidence |
| CharXiv | 68.5% | measured |
| CharXiv w/o tools | 76.1% | estimated ± 1.2 pp, high confidence |
| ExploitBench | 32.4% | estimated ± 6.5 pp, low confidence |
| ExploitGym | 1.9% | estimated ± 8.3 pp, low confidence |
| AI2D_TEST | 87.7% | measured |
| C-Eval | 92.2% | measured |
| Claw-Eval | 59.6% | measured |
| CountBench | 90.6% | measured |
| DeepPlanning | 26.4% | measured |
| DynaMath | 79.7% | measured |
| ERQA | 46.8% | measured |
| HMMT Feb 2025 | 92.9% | measured |
| HMMT Feb 2026 | 85.3% | measured |
| HMMT Nov 2025 | 93.3% | measured |
| INCLUDE | 86.2% | measured |
| MathVision | 74.3% | measured |
| MAXIFE | 79.2% | measured |
| MCP Atlas | 42.3% | measured |
| MCP-Tasks | 71.8% | measured |
| MLVU (M-Avg) | 81.7% | measured |
| MMAnswerBench | 84.0% | measured |
| MMLU-ProX | 85.7% | measured |
| MMLU-Redux | 96.6% | measured |
| MMMU-Pro | 70.6% | measured |
| NL2Repo | 43.2% | measured |
| NOVA-63 | 56.7% | measured |
| OmniDocBench 1.5 | 87.7% | measured |
| PolyMath | 79.0% | measured |
| PostTrainBench v1 | 17.1% | measured |
| QwenClawBench | 52.3% | measured |
| RealWorldQA | 77.0% | measured |
| ScreenSpot Pro | 45.7% | measured |
| SimpleVQA | 65.7% | measured |
| SuperGPQA | 70.6% | measured |
| Terminal-Bench 2.0 | 59.3% | measured |
| Toolathlon | 43.5% | measured |
| V* | 67.0% | measured |
| Video-MME (w/o subtitle) | 81.4% | measured |
| Video-MME (with subtitle) | 77.6% | measured |
| VideoMMMU | 84.4% | measured |
| We-Math | 70.0% | measured |
| HealthBench Hard | 0.0% | estimated ± 5.1 pp, low confidence |
| LiveCodeBench Pro | 81.4% | estimated ± 6.0 pp, low confidence |
| MedXpertQA (MM) | 52.4% | estimated ± 2.4 pp, high confidence |
| MMLU-Pro (Arcee) | 78.3% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 80.1% | estimated ± 2.7 pp, medium confidence |
| ResearchClawBench | 16.3% | estimated ± 1.3 pp, high confidence |
| SWE-Rebench | 61.8% | estimated ± 3.6 pp, medium confidence |
| PostTrainBench | 14.8% | estimated ± 1.2 pp, low confidence |
| DAYJOB | 1.4% | estimated ± 2.7 pp, high confidence |
| DAYJOB Healthcare | 0.0% | estimated ± 4.0 pp, high confidence |
| Finance Agent v2 | 55.5% | estimated ± 2.4 pp, medium confidence |
| BioMysteryBench (human-solvable) | 81.7% | estimated ± 0.6 pp, low confidence |
| FrontierCode 1.1 Extended | 55.0% | estimated ± 3.0 pp, low confidence |
| HealthBench (length-adjusted) | 54.1% | estimated ± 3.7 pp, medium confidence |
| HealthBench Professional | 56.8% | estimated ± 4.2 pp, low confidence |
| HLE-Verified | 35.7% | estimated ± 7.1 pp, low confidence |
| LABBench2 | 80.7% | estimated ± 1.8 pp, low confidence |
| BlenderWorld (text-only) | 70.2% | estimated ± 2.8 pp, medium confidence |
| Cookie-Bench | 77.3% | estimated ± 1.3 pp, medium confidence |
| STEMBench | 2.5% | estimated ± 0.2 pp, medium confidence |
| SimpleQA | 46.8% | estimated ± 4.3 pp, low confidence |
| LiveCodeBench | 79.8% | estimated ± 14.6 pp, low confidence |
| Agents' Last Exam | 30.7% | estimated ± 9.7 pp, medium confidence |
| BabyVision w/ Python | 88.8% | estimated ± 3.8 pp, low confidence |
| OpenHarmony Bench v1.0 | 51.2% | estimated ± 2.7 pp, high confidence |
| SEC-Bench Pro | 29.6% | estimated ± 7.5 pp, low confidence |
| IMOAnswerBench | 80.6% | estimated ± 5.6 pp, medium confidence |
| BabyVision | 22.7% | estimated ± 3.7 pp, medium confidence |
| MMVU | 72.5% | estimated ± 2.7 pp, medium confidence |
| PerceptionBench | 55.6% | estimated ± 4.9 pp, low confidence |
| MRCRv2 | 76.6% | estimated ± 11.3 pp, low confidence |
| CC-OCR | 76.1% | estimated ± 3.7 pp, medium confidence |
| MMLongBench-Doc | 61.0% | estimated ± 2.6 pp, low confidence |
| RefCOCO (avg) | 88.4% | estimated ± 3.1 pp, medium confidence |
| OCRBench V2 | 70.6% | estimated ± 2.1 pp, medium confidence |
| BioBasicsBench | 34.1% | estimated ± 3.8 pp, medium confidence |
| LifeVal | 45.6% | estimated ± 5.0 pp, low confidence |
| MMMU-Pro w/ Python | 74.8% | estimated ± 2.0 pp, high confidence |
| ExploitBench Internal Port | 0.0% | estimated ± 1.4 pp, low confidence |
| PinchBench | 94.5% | estimated ± 5.5 pp, low confidence |