xAI · model
Grok 4 benchmark scores
As of 2026-11-21, Grok 4 (xAI) has measured scores on 20 benchmarks and estimated scores on 89 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 87.7% | measured |
| AA-HLE | 26.7% | measured |
| AA-SciCode | 42.7% | estimated ± 4.2 pp, high confidence |
| Artificial Analysis Intelligence Index | 22.5% | measured |
| CritPt | 2.0% | measured |
| GDPval-AA | 17.9% | estimated ± 8.8 pp, medium confidence |
| GPQA Diamond | 89.8% | estimated ± 1.8 pp, high confidence |
| HLE | 28.7% | estimated ± 8.5 pp, medium confidence |
| IFBench | 57.1% | estimated ± 6.6 pp, medium confidence |
| AA-LCR | 68.0% | measured |
| LiveCodeBench v6 | 87.9% | estimated ± 2.2 pp, medium confidence |
| AA-Omniscience Accuracy | 40.5% | measured |
| SciCode | 43.1% | estimated ± 2.9 pp, low confidence |
| BrowseComp | 63.3% | estimated ± 6.6 pp, medium confidence |
| HLE w/ tools | 25.1% | estimated ± 5.5 pp, low confidence |
| IFEval | 91.6% | estimated ± 2.8 pp, high confidence |
| LongBench v2 | 57.5% | estimated ± 5.5 pp, medium confidence |
| τ²-bench results | 74.9% | measured |
| AA Coding Index | 34.7% | estimated ± 5.6 pp, medium confidence |
| AA-MMMU-Pro | 68.8% | measured |
| APEX-Agents-AA | 12.2% | estimated ± 11.2 pp, low confidence |
| BioMysteryBench (human-difficult) | 25.6% | estimated ± 6.8 pp, low confidence |
| SWE-bench Verified | 70.5% | estimated ± 4.4 pp, high confidence |
| BFCL v4 | 58.6% | estimated ± 8.5 pp, low confidence |
| CyberGym | 34.8% | estimated ± 9.3 pp, low confidence |
| JobBench | 18.7% | estimated ± 10.3 pp, low confidence |
| SWE-bench Pro | 57.2% | estimated ± 4.7 pp, high confidence |
| WideResearch | 63.4% | estimated ± 5.9 pp, low confidence |
| AA Agentic Index | 12.3% | estimated ± 6.8 pp, medium confidence |
| MMLU-Pro | 85.0% | estimated ± 4.8 pp, high confidence |
| AA-IFBench | 53.7% | measured |
| FailureSensorIQ | 60.4% | estimated ± 1.8 pp, low confidence |
| GPQA | 89.7% | estimated ± 4.5 pp, high confidence |
| MMLU | 90.5% | estimated ± 0.9 pp, medium confidence |
| Vals GPQA Diamond | 88.1% | measured |
| Vals LiveCodeBench | 83.2% | measured |
| Vals MMLU-Pro | 85.3% | measured |
| MMMLU | 87.2% | estimated ± 0.7 pp, medium confidence |
| MMMU | 79.3% | estimated ± 2.7 pp, high confidence |
| AA MMLU-Pro | 86.6% | measured |
| Gert Labs | 42.3% | measured |
| ARC-AGI-1 | 83.7% | estimated ± 6.5 pp, medium confidence |
| ARC-AGI-2 | 51.2% | estimated ± 13.8 pp, low confidence |
| CursorBench 3.1 | 31.3% | estimated ± 5.8 pp, low confidence |
| CursorBench 3.2 | 59.1% | estimated ± 7.2 pp, low confidence |
| FrontierCode 1.1 Main | 9.8% | estimated ± 5.2 pp, low confidence |
| Mystery Game Puzzles | 6.9% | estimated ± 2.2 pp, low confidence |
| OfficeQA Pro | 50.3% | estimated ± 6.8 pp, low confidence |
| OSWorld-Verified | 56.8% | estimated ± 5.5 pp, medium confidence |
| ParseBench-1 | 52.7% | estimated ± 6.2 pp, medium confidence |
| Vals SWE-bench | 57.8% | measured |
| VulcanBench v3 | 82.5% | estimated ± 5.6 pp, low confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 52.7% | estimated ± 4.3 pp, low confidence |
| HealthBench Professional (raw, Artificial Analysis) | 57.5% | estimated ± 1.1 pp, low confidence |
| AA ITBench | 36.5% | estimated ± 2.8 pp, high confidence |
| Chartography (tools) | 76.1% | estimated ± 4.1 pp, low confidence |
| CursorBench 4.0 | 40.0% | estimated ± 4.5 pp, high confidence |
| HLE w/o tools | 31.3% | estimated ± 4.9 pp, high confidence |
| OSWorld 2.0 | 0.0% | estimated ± 5.9 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 8.6 pp, low confidence |
| ARC-AGI-3 Provider Adapter | 2.7% | estimated ± 7.3 pp, low confidence |
| SkillsBench official leaderboard (BenchFlow) | 30.2% | estimated ± 9.5 pp, low confidence |
| BinaryAudit | 8.8% | measured |
| CompileBench | 71.1% | measured |
| OTelBench | 4.3% | measured |
| Vibe Code Bench | 27.7% | estimated ± 6.5 pp, low confidence |
| ParseBench-2 | 56.4% | estimated ± 6.2 pp, medium confidence |
| CharXiv | 76.8% | estimated ± 6.3 pp, low confidence |
| CharXiv w/o tools | 76.1% | estimated ± 2.5 pp, medium confidence |
| AI2D_TEST | 88.1% | estimated ± 2.0 pp, medium confidence |
| C-Eval | 92.3% | estimated ± 1.0 pp, medium confidence |
| Claw-Eval | 56.5% | estimated ± 6.8 pp, low confidence |
| CountBench | 90.7% | estimated ± 2.9 pp, low confidence |
| DynaMath | 52.6% | estimated ± 3.0 pp, low confidence |
| ERQA | 42.3% | estimated ± 3.1 pp, medium confidence |
| MathVision | 75.8% | estimated ± 3.1 pp, medium confidence |
| MCP Atlas | 38.7% | estimated ± 14.6 pp, low confidence |
| MMLU-Redux | 96.0% | estimated ± 0.8 pp, medium confidence |
| MMMU-Pro | 70.3% | estimated ± 1.5 pp, high confidence |
| NL2Repo | 36.8% | estimated ± 4.7 pp, medium confidence |
| PostTrainBench v1 | 0.0% | estimated ± 3.6 pp, medium confidence |
| RealWorldQA | 78.9% | estimated ± 2.9 pp, low confidence |
| ScreenSpot Pro | 70.5% | estimated ± 10.8 pp, low confidence |
| SuperGPQA | 82.1% | estimated ± 9.3 pp, medium confidence |
| Toolathlon | 36.6% | estimated ± 7.2 pp, medium confidence |
| V* | 34.5% | estimated ± 7.5 pp, low confidence |
| Video-MME (w/o subtitle) | 79.0% | estimated ± 1.7 pp, medium confidence |
| Video-MME (with subtitle) | 80.9% | estimated ± 3.7 pp, low confidence |
| AA LiveCodeBench | 81.9% | measured |
| HealthBench Hard | 15.3% | estimated ± 8.6 pp, low confidence |
| MedXpertQA (MM) | 42.5% | estimated ± 6.5 pp, low confidence |
| MMLU-Pro (Arcee) | 84.6% | estimated ± 4.0 pp, medium confidence |
| React Native Evals | 72.6% | measured |
| ResearchClawBench | 14.7% | estimated ± 2.2 pp, medium confidence |
| BioMysteryBench (human-solvable) | 81.7% | estimated ± 0.6 pp, low confidence |
| HealthBench (raw) | 35.4% | estimated ± 6.0 pp, low confidence |
| HealthBench (length-adjusted) | 48.5% | estimated ± 3.4 pp, low confidence |
| HealthBench Professional | 56.7% | estimated ± 4.2 pp, low confidence |
| HealthBench Professional (raw) | 46.8% | estimated ± 4.5 pp, medium confidence |
| HLE-Verified | 15.3% | estimated ± 2.0 pp, low confidence |
| LABBench2 | 80.5% | estimated ± 1.8 pp, low confidence |
| STEMBench | 2.5% | estimated ± 0.2 pp, medium confidence |
| SimpleQA | 46.3% | estimated ± 4.3 pp, low confidence |
| BabyVision w/ Python | 83.4% | estimated ± 4.0 pp, low confidence |
| OpenHarmony Bench v1.0 | 47.3% | estimated ± 3.2 pp, medium confidence |
| MRCRv2 | 64.2% | estimated ± 11.3 pp, low confidence |
| MMLongBench-Doc | 60.8% | estimated ± 2.6 pp, low confidence |
| BioBasicsBench | 33.9% | estimated ± 2.8 pp, low confidence |
| MMMU-Pro w/ Python | 72.6% | estimated ± 2.0 pp, high confidence |