Anthropic · model
Claude Fable 5 benchmark scores
As of 2026-11-22, Claude Fable 5 (Anthropic) has measured scores on 51 benchmarks and estimated scores on 121 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 92.6% | measured |
| AA-HLE | 55.5% | measured |
| AA-SciCode | 61.0% | measured |
| AIME26 | 98.0% | estimated ± 1.4 pp, medium confidence |
| Artificial Analysis Intelligence Index | 49.6% | measured |
| CritPt | 28.6% | measured |
| GDPval-AA | 55.5% | measured |
| GPQA Diamond | 96.4% | estimated ± 1.4 pp, high confidence |
| HLE | 61.4% | estimated ± 6.1 pp, medium confidence |
| IFBench | 62.4% | estimated ± 6.6 pp, medium confidence |
| AA-LCR | 82.3% | measured |
| LiveCodeBench v6 | 90.8% | estimated ± 2.2 pp, low confidence |
| AA-Omniscience Accuracy | 65.4% | measured |
| SciCode | 56.4% | estimated ± 2.9 pp, low confidence |
| Terminal-Bench 2.1 | 84.3% | measured |
| BrowseComp | 91.2% | estimated ± 2.9 pp, high confidence |
| HLE w/ tools | 60.6% | estimated ± 5.3 pp, medium confidence |
| IFEval | 92.4% | estimated ± 2.8 pp, high confidence |
| LongBench v2 | 67.7% | estimated ± 5.5 pp, medium confidence |
| τ²-bench results | 98.5% | measured |
| AA Coding Index | 76.5% | measured |
| AA-MMMU-Pro | 78.6% | estimated ± 2.3 pp, high confidence |
| APEX-Agents | 49.7% | estimated ± 9.4 pp, low confidence |
| APEX-Agents-AA | 46.0% | estimated ± 7.8 pp, low confidence |
| BioMysteryBench (human-difficult) | 49.5% | estimated ± 6.8 pp, low confidence |
| DeepSearchQA | 93.5% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 95.0% | measured |
| AutomationBench | 24.5% | estimated ± 7.1 pp, low confidence |
| BFCL v4 | 77.0% | estimated ± 2.7 pp, medium confidence |
| CyberGym | 79.9% | estimated ± 3.1 pp, low confidence |
| JobBench | 60.2% | estimated ± 8.7 pp, medium confidence |
| SWE-bench Pro | 80.0% | measured |
| WideResearch | 81.4% | estimated ± 2.1 pp, medium confidence |
| AA Agentic Index | 51.0% | measured |
| MMLU-Pro | 88.5% | estimated ± 4.8 pp, medium confidence |
| AA-IFBench | 63.5% | measured |
| FailureSensorIQ | 68.2% | estimated ± 1.7 pp, high confidence |
| GPQA | 95.0% | estimated ± 1.6 pp, high confidence |
| MMLU | 97.3% | estimated ± 1.2 pp, medium confidence |
| Vals GPQA Diamond | 93.2% | measured |
| Vals LiveCodeBench | 89.8% | measured |
| Vals MMLU-Pro | 91.5% | measured |
| MMMLU | 89.7% | estimated ± 1.7 pp, medium confidence |
| Chess Puzzles | 47.0% | measured |
| AA MMLU-Pro | 91.0% | estimated ± 0.7 pp, medium confidence |
| Gert Labs | 72.7% | measured |
| AA-AnalystAgent | 48.8% | measured |
| AA EnterpriseOps-Gym | 51.1% | measured |
| AA Harvey LAB v1.0 | 93.6% | measured |
| AA Tau3 Banking | 38.1% | measured |
| AA Terminal-Bench 2.1 | 84.6% | measured |
| ApprenticeBench | 34.0% | measured |
| ARC-AGI-1 | 98.5% | measured |
| ARC-AGI-2 | 89.2% | measured |
| BabaIsBench | 100.0% | measured |
| CursorBench 3.1 | 70.6% | measured |
| CursorBench 3.2 | 70.5% | measured |
| EBR-bench | 39.5% | measured |
| Terminal-Bench 3.0 | 34.1% | measured |
| FrontierCode 1.1 Main | 53.5% | measured |
| FrontierMath v2 (Tier 4) | 90.2% | measured |
| FrontierMath v2 (Tiers 1-3) | 87.0% | measured |
| FrontierSWE v2 | 47.0% | measured |
| Furniture Assembly | 35.8% | measured |
| MaxxBench | 63.5% | measured |
| MazeBench | 11.0% | measured |
| MirrorCode | 70.0% | measured |
| Mystery Game Puzzles | 52.0% | measured |
| NYT Connections | 92.7% | measured |
| OfficeQA Pro | 57.9% | measured |
| OSWorld-Verified | 85.0% | measured |
| ParseBench-1 | 69.8% | measured |
| Terminal-Bench 4.1 | 44.9% | measured |
| Terminal-Bench Hard | 62.9% | measured |
| Vals SWE-bench | 95.0% | measured |
| Terminal-Bench 2.1 (Vals AI) | 80.5% | measured |
| VulcanBench v3 | 89.5% | measured |
| CWE-Bench-AA | 60.5% | estimated ± 8.1 pp, medium confidence |
| AA Cyber Index | 29.6% | estimated ± 9.2 pp, low confidence |
| DeepsecBench-AA | 29.7% | estimated ± 8.2 pp, medium confidence |
| Harvey LAB-AA v1.1 | 7.4% | estimated ± 4.0 pp, low confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 61.0% | estimated ± 4.3 pp, medium confidence |
| HealthBench Professional (raw, Artificial Analysis) | 69.4% | estimated ± 0.8 pp, medium confidence |
| AA ITBench | 51.6% | estimated ± 3.2 pp, high confidence |
| Bug Hunt Bench | 25.4% | estimated ± 7.7 pp, medium confidence |
| Chartography (tools) | 79.5% | estimated ± 4.3 pp, low confidence |
| CursorBench 4.0 | 44.4% | estimated ± 2.0 pp, high confidence |
| CWE-bench v1 | 59.5% | estimated ± 5.1 pp, medium confidence |
| DeepSWE | 72.3% | estimated ± 4.3 pp, high confidence |
| DRACO | 87.9% | estimated ± 0.7 pp, low confidence |
| FrontierCode 2.0 | 29.0% | estimated ± 4.8 pp, high confidence |
| FrontierPhysics | 1.8% | estimated ± 4.4 pp, high confidence |
| HLE w/o tools | 54.6% | estimated ± 4.9 pp, high confidence |
| OSWorld 2.0 | 19.9% | estimated ± 3.6 pp, high confidence |
| OSWorld 2.1 | 77.6% | estimated ± 3.6 pp, high confidence |
| OSWorld 2.1 Offline | 71.9% | estimated ± 1.9 pp, medium confidence |
| PostTrainBench v1.1 | 34.4% | estimated ± 3.7 pp, high confidence |
| ProgramBench | 89.1% | estimated ± 6.0 pp, low confidence |
| SpireClimb | 5.8% | estimated ± 7.2 pp, medium confidence |
| StealthMark | 45.8% | estimated ± 6.5 pp, medium confidence |
| SWE-bench Multimodal | 56.7% | estimated ± 5.6 pp, low confidence |
| Terminal-Bench 4.0 | 48.1% | estimated ± 4.5 pp, high confidence |
| Terminal-Bench-Science 0.1 | 33.8% | estimated ± 10.7 pp, low confidence |
| Toolathlon-Verified | 74.7% | estimated ± 2.3 pp, medium confidence |
| VulcanBench Frontier v4 | 90.8% | estimated ± 2.3 pp, medium confidence |
| AA AutomationBench | 66.6% | estimated ± 9.9 pp, low confidence |
| GDP.pdf | 25.6% | estimated ± 4.2 pp, high confidence |
| AA Terminal-Bench 4.0 | 44.5% | estimated ± 6.2 pp, medium confidence |
| ARC-AGI-3 | 11.4% | estimated ± 8.6 pp, low confidence |
| ARC-AGI-3 Provider Adapter | 97.1% | estimated ± 7.3 pp, medium confidence |
| cursorBench41 | 53.7% | estimated ± 3.7 pp, high confidence |
| CWE-bench | 60.1% | estimated ± 4.6 pp, high confidence |
| MKA | 46.6% | estimated ± 4.6 pp, high confidence |
| SkillsBench official leaderboard (BenchFlow) | 57.1% | estimated ± 6.1 pp, medium confidence |
| BinaryAudit | 35.8% | estimated ± 12.1 pp, low confidence |
| OTelBench | 22.5% | estimated ± 6.6 pp, medium confidence |
| Vibe Code Bench | 81.0% | estimated ± 5.6 pp, low confidence |
| FACTS Multimodal | 51.8% | estimated ± 4.0 pp, high confidence |
| Hermes Bench | 72.0% | estimated ± 2.9 pp, high confidence |
| Hermes Index | 58.8% | estimated ± 3.1 pp, high confidence |
| ParseBench-2 | 75.2% | estimated ± 4.3 pp, high confidence |
| CharXiv | 86.9% | estimated ± 5.1 pp, low confidence |
| CharXiv w/o tools | 84.7% | estimated ± 1.8 pp, medium confidence |
| ExploitBench | 32.4% | estimated ± 6.5 pp, medium confidence |
| ExploitGym | 30.8% | estimated ± 2.7 pp, high confidence |
| C-Eval | 94.4% | estimated ± 1.2 pp, low confidence |
| Claw-Eval | 71.9% | estimated ± 6.2 pp, low confidence |
| HMMT Feb 2026 | 100.0% | estimated ± 3.1 pp, medium confidence |
| HMMT Nov 2025 | 96.0% | estimated ± 1.5 pp, low confidence |
| MCP Atlas | 84.7% | estimated ± 5.8 pp, low confidence |
| MMAnswerBench | 96.0% | estimated ± 2.4 pp, low confidence |
| MMLU-Redux | 95.8% | estimated ± 3.3 pp, medium confidence |
| MMMU-Pro | 81.6% | estimated ± 6.9 pp, medium confidence |
| NL2Repo | 58.9% | estimated ± 4.7 pp, high confidence |
| PostTrainBench v1 | 42.9% | estimated ± 3.6 pp, medium confidence |
| ScreenSpot Pro | 89.2% | estimated ± 10.1 pp, low confidence |
| SuperGPQA | 100.0% | estimated ± 9.3 pp, low confidence |
| Terminal-Bench 2.0 | 81.9% | estimated ± 5.4 pp, low confidence |
| Toolathlon | 58.9% | estimated ± 1.7 pp, low confidence |
| AA LiveCodeBench | 100.0% | estimated ± 2.6 pp, medium confidence |
| HealthBench Hard | 33.8% | estimated ± 8.6 pp, low confidence |
| LiveCodeBench Pro | 97.2% | estimated ± 6.0 pp, low confidence |
| React Native Evals | 86.4% | estimated ± 2.7 pp, low confidence |
| ResearchClawBench | 21.2% | estimated ± 1.9 pp, medium confidence |
| SWE-Rebench | 61.8% | estimated ± 3.6 pp, low confidence |
| PostTrainBench | 33.2% | estimated ± 2.4 pp, high confidence |
| DAYJOB | 1.4% | estimated ± 2.7 pp, high confidence |
| DAYJOB Healthcare | 0.0% | estimated ± 4.0 pp, high confidence |
| Finance Agent v2 | 61.2% | estimated ± 2.4 pp, medium confidence |
| BioMysteryBench (human-solvable) | 89.5% | estimated ± 0.6 pp, medium confidence |
| FrontierCode 1.1 Extended | 62.4% | estimated ± 1.6 pp, medium confidence |
| HealthBench (raw) | 61.8% | estimated ± 6.0 pp, medium confidence |
| HealthBench (length-adjusted) | 58.4% | estimated ± 2.5 pp, low confidence |
| HealthBench Professional | 60.5% | estimated ± 4.2 pp, medium confidence |
| HealthBench Professional (raw) | 71.7% | estimated ± 4.5 pp, high confidence |
| HLE-Verified | 54.2% | estimated ± 1.3 pp, medium confidence |
| LABBench2 | 86.1% | estimated ± 1.7 pp, medium confidence |
| ScarfBench | 68.9% | estimated ± 4.5 pp, low confidence |
| BlenderWorld (text-only) | 80.2% | estimated ± 2.8 pp, medium confidence |
| Cookie-Bench | 88.7% | estimated ± 1.3 pp, medium confidence |
| STEMBench | 2.7% | estimated ± 0.2 pp, medium confidence |
| SimpleQA | 34.9% | estimated ± 7.8 pp, low confidence |
| LiveCodeBench | 87.6% | estimated ± 14.6 pp, low confidence |
| Agents' Last Exam | 43.4% | estimated ± 9.7 pp, medium confidence |
| OpenHarmony Bench v1.0 | 58.9% | estimated ± 2.4 pp, medium confidence |
| SEC-Bench Pro | 60.3% | estimated ± 7.5 pp, medium confidence |
| MRCRv2 | 85.9% | estimated ± 11.3 pp, low confidence |
| MMLongBench-Doc | 61.5% | estimated ± 2.6 pp, low confidence |
| BioBasicsBench | 67.7% | estimated ± 2.8 pp, low confidence |
| LifeVal | 79.4% | estimated ± 2.3 pp, medium confidence |
| MMMU-Pro w/ Python | 83.1% | estimated ± 2.7 pp, medium confidence |
| ExploitBench Internal Port | 1.2% | estimated ± 1.4 pp, medium confidence |