Xiaomi · model
MiMo-V2.6-Pro benchmark scores
As of 2026-11-21, MiMo-V2.6-Pro (Xiaomi) has measured scores on 34 benchmarks and estimated scores on 104 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 91.8% | estimated ± 9.2 pp, medium confidence |
| AA-HLE | 49.4% | measured |
| AA-SciCode | 60.9% | measured |
| Artificial Analysis Intelligence Index | 46.3% | measured |
| CritPt | 26.6% | measured |
| GDPval-AA | 59.2% | measured |
| GPQA Diamond | 94.0% | estimated ± 2.4 pp, high confidence |
| HLE | 57.5% | estimated ± 8.5 pp, medium confidence |
| AA-LCR | 86.3% | measured |
| LiveCodeBench v6 | 100.0% | estimated ± 4.9 pp, medium confidence |
| AA-Omniscience Accuracy | 34.8% | measured |
| SciCode | 60.4% | estimated ± 3.1 pp, medium confidence |
| Terminal-Bench 2.1 | 89.9% | measured |
| BrowseComp | 86.6% | estimated ± 3.0 pp, medium confidence |
| HLE w/ tools | 63.8% | estimated ± 4.9 pp, low confidence |
| LongBench v2 | 71.5% | estimated ± 5.5 pp, low confidence |
| AA Coding Index | 76.7% | estimated ± 2.5 pp, high confidence |
| APEX-Agents | 68.4% | estimated ± 7.4 pp, low confidence |
| APEX-Agents-AA | 41.7% | estimated ± 3.6 pp, low confidence |
| BioMysteryBench (human-difficult) | 49.3% | estimated ± 6.8 pp, low confidence |
| DeepSearchQA | 87.7% | estimated ± 7.9 pp, low confidence |
| SWE-bench Verified | 89.0% | estimated ± 5.1 pp, medium confidence |
| AutomationBench | 53.1% | measured |
| BFCL v4 | 77.2% | estimated ± 2.7 pp, medium confidence |
| CyberGym | 94.0% | measured |
| JobBench | 62.0% | measured |
| SWE-bench Pro | 77.5% | estimated ± 6.3 pp, medium confidence |
| WideResearch | 83.9% | estimated ± 1.2 pp, low confidence |
| AA Agentic Index | 56.4% | estimated ± 3.1 pp, high confidence |
| MMLU-Pro | 85.5% | estimated ± 5.0 pp, medium confidence |
| GPQA | 92.8% | estimated ± 7.8 pp, medium confidence |
| MMLU | 90.5% | estimated ± 0.9 pp, medium confidence |
| Vals GPQA Diamond | 92.1% | estimated ± 5.7 pp, medium confidence |
| Vals LiveCodeBench | 89.0% | estimated ± 5.7 pp, low confidence |
| Vals MMLU-Pro | 86.3% | estimated ± 2.9 pp, high confidence |
| MMMLU | 91.4% | estimated ± 0.7 pp, low confidence |
| Chess Puzzles | 41.1% | estimated ± 8.0 pp, medium confidence |
| AA MMLU-Pro | 86.4% | estimated ± 0.8 pp, high confidence |
| Gert Labs | 73.1% | estimated ± 5.5 pp, low confidence |
| AA-AnalystAgent | 37.5% | measured |
| AA EnterpriseOps-Gym | 46.7% | estimated ± 1.2 pp, medium confidence |
| AA Tau3 Banking | 100.0% | estimated ± 2.4 pp, low confidence |
| AA Terminal-Bench 2.1 | 82.1% | estimated ± 2.5 pp, high confidence |
| ApprenticeBench | 35.4% | estimated ± 4.0 pp, medium confidence |
| ARC-AGI-1 | 94.7% | estimated ± 7.8 pp, low confidence |
| ARC-AGI-2 | 81.7% | estimated ± 14.4 pp, low confidence |
| CursorBench 3.1 | 70.9% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 27.5% | estimated ± 3.6 pp, low confidence |
| EBR-bench | 47.1% | estimated ± 9.9 pp, medium confidence |
| Terminal-Bench 3.0 | 24.8% | estimated ± 3.2 pp, high confidence |
| FrontierCode 1.1 Main | 50.8% | estimated ± 3.5 pp, low confidence |
| FrontierSWE v2 | 31.0% | estimated ± 11.1 pp, low confidence |
| Furniture Assembly | 46.8% | estimated ± 6.5 pp, medium confidence |
| MaxxBench | 55.2% | estimated ± 5.7 pp, medium confidence |
| MazeBench | 2.0% | estimated ± 3.7 pp, high confidence |
| MirrorCode | 34.5% | estimated ± 14.9 pp, low confidence |
| Mystery Game Puzzles | 27.0% | estimated ± 7.5 pp, medium confidence |
| NYT Connections | 100.0% | estimated ± 8.9 pp, low confidence |
| OSWorld-Verified | 82.0% | measured |
| Terminal-Bench 4.1 | 27.0% | measured |
| Terminal-Bench Hard | 59.9% | estimated ± 2.8 pp, medium confidence |
| Vals SWE-bench | 94.2% | estimated ± 8.2 pp, medium confidence |
| Terminal-Bench 2.1 (Vals AI) | 67.8% | measured |
| VulcanBench v3 | 78.2% | estimated ± 3.9 pp, medium confidence |
| CWE-Bench-AA | 63.3% | measured |
| AA Cyber Index | 56.1% | measured |
| DeepsecBench-AA | 26.2% | measured |
| Harvey LAB-AA v1.1 | 0.8% | estimated ± 3.5 pp, low confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 60.0% | estimated ± 4.3 pp, medium confidence |
| HealthBench Professional (raw, Artificial Analysis) | 68.7% | estimated ± 0.8 pp, medium confidence |
| AA ITBench | 44.3% | estimated ± 4.7 pp, low confidence |
| Bug Hunt Bench | 22.7% | measured |
| CursorBench 4.0 | 43.3% | estimated ± 4.7 pp, high confidence |
| CWE-bench v1 | 55.0% | estimated ± 4.5 pp, medium confidence |
| DeepSWE | 71.9% | measured |
| DRACO | 86.2% | estimated ± 2.0 pp, medium confidence |
| FrontierCode 2.0 | 23.8% | estimated ± 6.7 pp, medium confidence |
| FrontierPhysics | 1.2% | measured |
| HLE w/o tools | 53.0% | estimated ± 4.9 pp, high confidence |
| OSWorld 2.0 | 16.7% | estimated ± 3.6 pp, high confidence |
| OSWorld 2.1 | 73.5% | estimated ± 6.0 pp, medium confidence |
| OSWorld 2.1 Offline | 78.0% | estimated ± 1.9 pp, medium confidence |
| PostTrainBench v1.1 | 39.9% | estimated ± 4.5 pp, high confidence |
| ProgramBench | 26.5% | measured |
| SpireClimb | 9.3% | estimated ± 5.4 pp, low confidence |
| StealthMark | 47.9% | measured |
| Terminal-Bench 4.0 | 34.9% | measured |
| Terminal-Bench-Science 0.1 | 4.3% | estimated ± 10.7 pp, low confidence |
| Toolathlon-Verified | 76.9% | measured |
| VulcanBench Frontier v4 | 88.5% | estimated ± 1.2 pp, low confidence |
| AA AutomationBench | 58.6% | measured |
| GDP.pdf | 19.2% | measured |
| AA Terminal-Bench 4.0 | 34.8% | measured |
| ARC-AGI-3 | 7.5% | estimated ± 9.8 pp, medium confidence |
| ARC-AGI-3 Provider Adapter | 62.3% | estimated ± 7.3 pp, medium confidence |
| cursorBench41 | 52.9% | estimated ± 3.7 pp, high confidence |
| CWE-bench | 48.1% | estimated ± 4.4 pp, high confidence |
| MKA | 37.3% | measured |
| FACTS Multimodal | 40.0% | estimated ± 2.7 pp, medium confidence |
| Hermes Bench | 72.7% | estimated ± 1.5 pp, medium confidence |
| Hermes Index | 37.4% | estimated ± 2.7 pp, medium confidence |
| ExploitBench | 47.9% | measured |
| ExploitGym | 17.8% | measured |
| C-Eval | 93.3% | estimated ± 1.5 pp, low confidence |
| Claw-Eval | 77.1% | estimated ± 4.5 pp, medium confidence |
| MCP Atlas | 83.8% | estimated ± 7.5 pp, medium confidence |
| MMLU-Redux | 95.3% | estimated ± 0.8 pp, medium confidence |
| NL2Repo | 61.6% | estimated ± 6.0 pp, medium confidence |
| SuperGPQA | 73.2% | estimated ± 9.3 pp, medium confidence |
| Terminal-Bench 2.0 | 71.6% | estimated ± 5.8 pp, medium confidence |
| Toolathlon | 55.9% | estimated ± 6.3 pp, low confidence |
| HealthBench Hard | 33.6% | estimated ± 8.6 pp, low confidence |
| React Native Evals | 81.3% | estimated ± 4.7 pp, low confidence |
| ResearchClawBench | 19.7% | estimated ± 2.1 pp, medium confidence |
| PostTrainBench | 33.1% | estimated ± 5.0 pp, medium confidence |
| DAYJOB | 1.8% | estimated ± 1.1 pp, medium confidence |
| DAYJOB Healthcare | 1.5% | estimated ± 1.5 pp, medium confidence |
| BioMysteryBench (human-solvable) | 89.5% | estimated ± 0.6 pp, medium confidence |
| HealthBench (raw) | 56.9% | estimated ± 6.0 pp, medium confidence |
| HealthBench (length-adjusted) | 57.2% | estimated ± 3.4 pp, medium confidence |
| HealthBench Professional | 59.4% | estimated ± 4.2 pp, medium confidence |
| HealthBench Professional (raw) | 67.1% | estimated ± 4.5 pp, high confidence |
| HLE-Verified | 54.4% | estimated ± 6.8 pp, low confidence |
| LABBench2 | 83.8% | estimated ± 2.4 pp, medium confidence |
| ScarfBench | 67.5% | estimated ± 3.8 pp, medium confidence |
| STEMBench | 2.2% | measured |
| SimpleQA | 38.4% | estimated ± 4.3 pp, low confidence |
| Agents' Last Exam | 31.6% | measured |
| OpenHarmony Bench v1.0 | 58.9% | estimated ± 2.4 pp, medium confidence |
| OpenHarmony Bench v1.2 | 63.8% | estimated ± 11.0 pp, low confidence |
| SEC-Bench Pro | 66.3% | measured |
| MRCRv2 | 86.0% | estimated ± 11.3 pp, low confidence |
| MMLongBench-Doc | 61.7% | estimated ± 2.6 pp, low confidence |
| SWE-Marathon | 44.5% | estimated ± 2.1 pp, medium confidence |
| BioBasicsBench | 44.9% | estimated ± 2.8 pp, medium confidence |
| LifeVal | 71.5% | estimated ± 2.3 pp, low confidence |
| ExploitBench Internal Port | 3.6% | estimated ± 6.5 pp, medium confidence |
| SWE-Atlas Codebase QnA | 61.0% | estimated ± 4.8 pp, high confidence |