xAI · model
Grok 4.7 benchmark scores
As of 2026-10-07, Grok 4.7 (xAI) has measured scores on 19 benchmarks and estimated scores on 72 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 91.5% | estimated ± 2.1 pp, medium confidence |
| AA-HLE | 43.1% | measured |
| AA-SciCode | 57.4% | measured |
| Artificial Analysis Intelligence Index | 46.5% | measured |
| CritPt | 17.7% | measured |
| GDPval-AA | 60.8% | measured |
| GPQA Diamond | 92.7% | estimated ± 1.0 pp, medium confidence |
| HLE | 50.2% | estimated ± 9.5 pp, medium confidence |
| AA-LCR | 76.7% | measured |
| LiveCodeBench v6 | 98.0% | estimated ± 3.3 pp, medium confidence |
| AA-Omniscience Accuracy | 47.4% | measured |
| SciCode | 56.9% | estimated ± 3.2 pp, high confidence |
| Terminal-Bench 2.1 | 87.6% | estimated ± 3.9 pp, high confidence |
| BrowseComp | 79.1% | estimated ± 4.7 pp, low confidence |
| HLE w/ tools | 65.2% | estimated ± 4.6 pp, medium confidence |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, medium confidence |
| τ²-bench results | 99.9% | estimated ± 4.8 pp, low confidence |
| AA Coding Index | 79.0% | estimated ± 1.0 pp, high confidence |
| APEX-Agents | 47.9% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 33.7% | estimated ± 3.2 pp, low confidence |
| BioMysteryBench (human-difficult) | 44.5% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 90.5% | estimated ± 8.4 pp, medium confidence |
| SWE-bench Verified | 91.3% | estimated ± 5.2 pp, low confidence |
| AutomationBench | 44.6% | estimated ± 7.6 pp, low confidence |
| BFCL v4 | 78.2% | estimated ± 2.6 pp, medium confidence |
| CyberGym | 92.1% | estimated ± 6.2 pp, low confidence |
| JobBench | 65.4% | estimated ± 8.6 pp, low confidence |
| SWE-bench Pro | 74.5% | estimated ± 4.3 pp, high confidence |
| WideResearch | 100.0% | estimated ± 9.1 pp, low confidence |
| AA Agentic Index | 55.1% | estimated ± 2.8 pp, high confidence |
| MMLU-Pro | 85.7% | estimated ± 3.5 pp, high confidence |
| GPQA | 92.7% | estimated ± 1.0 pp, medium confidence |
| Gert Labs | 81.3% | estimated ± 8.9 pp, low confidence |
| AA-AnalystAgent | 46.7% | estimated ± 6.5 pp, medium confidence |
| AA EnterpriseOps-Gym | 45.0% | estimated ± 4.4 pp, high confidence |
| AA Harvey LAB | 19.6% | measured |
| ARC-AGI-1 | 94.3% | estimated ± 5.4 pp, medium confidence |
| ARC-AGI-2 | 79.8% | estimated ± 14.2 pp, low confidence |
| CursorBench 3.1 | 62.2% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 70.6% | estimated ± 2.2 pp, high confidence |
| Terminal-Bench 3.0 | 17.2% | estimated ± 6.9 pp, medium confidence |
| FrontierCode 1.1 Main | 49.9% | estimated ± 3.7 pp, low confidence |
| FrontierSWE v2 | 29.5% | measured |
| OSWorld-Verified | 87.5% | estimated ± 8.7 pp, low confidence |
| Vals GPQA Diamond | 91.1% | estimated ± 1.5 pp, medium confidence |
| Vals LiveCodeBench | 88.6% | estimated ± 2.7 pp, low confidence |
| Vals MMLU-Pro | 86.8% | estimated ± 1.5 pp, medium confidence |
| Vals SWE-bench | 92.4% | estimated ± 4.9 pp, high confidence |
| Terminal-Bench 2.1 (Vals AI) | 73.4% | measured |
| VulcanBench v3 | 87.0% | estimated ± 0.7 pp, medium confidence |
| AA AutomationBench | 65.6% | measured |
| GDP.pdf | 20.0% | measured |
| AA ITBench | 42.1% | measured |
| AA Tau3 Banking | 48.2% | estimated ± 3.0 pp, low confidence |
| AA Terminal-Bench 2.1 | 86.4% | estimated ± 2.6 pp, high confidence |
| AA Terminal-Bench 4.0 | 25.8% | measured |
| CursorBench 4.0 | 46.3% | measured |
| CWE-bench v1 | 68.0% | measured |
| DeepSWE | 71.0% | measured |
| HLE w/o tools | 42.2% | estimated ± 5.7 pp, medium confidence |
| OSWorld 2.0 | 58.9% | estimated ± 13.3 pp, low confidence |
| PostTrainBench v1.1 | 33.4% | estimated ± 4.3 pp, medium confidence |
| Terminal-Bench 4.0 | 38.0% | measured |
| Terminal-Bench-Science 0.1 | 3.8% | estimated ± 4.1 pp, low confidence |
| Toolathlon-Verified | 76.4% | estimated ± 2.2 pp, medium confidence |
| ExploitGym | 15.8% | estimated ± 5.0 pp, low confidence |
| C-Eval | 96.3% | estimated ± 1.0 pp, low confidence |
| MCP Atlas | 86.7% | estimated ± 2.7 pp, low confidence |
| MMLU-Redux | 97.2% | estimated ± 1.1 pp, low confidence |
| NL2Repo | 61.6% | estimated ± 6.5 pp, medium confidence |
| QwenClawBench | 100.0% | estimated ± 6.0 pp, low confidence |
| SuperGPQA | 84.7% | estimated ± 7.5 pp, low confidence |
| Terminal-Bench 2.0 | 75.7% | estimated ± 5.9 pp, medium confidence |
| Toolathlon | 55.9% | estimated ± 6.3 pp, low confidence |
| HealthBench Hard | 34.6% | estimated ± 7.7 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 79.8% | estimated ± 4.7 pp, medium confidence |
| ResearchClawBench | 20.2% | estimated ± 2.2 pp, low confidence |
| ARC-AGI-3 | 1.7% | estimated ± 3.4 pp, high confidence |
| BioMysteryBench (human-solvable) | 88.7% | estimated ± 0.6 pp, medium confidence |
| DRACO | 86.5% | estimated ± 3.0 pp, medium confidence |
| FrontierCode 1.1 Extended | 61.0% | estimated ± 3.0 pp, medium confidence |
| HealthBench (raw) | 51.9% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 49.4% | estimated ± 2.2 pp, low confidence |
| HealthBench Professional | 56.7% | measured |
| HealthBench Professional (raw) | 62.1% | estimated ± 5.0 pp, medium confidence |
| HLE-Verified | 42.5% | estimated ± 6.0 pp, low confidence |
| LABBench2 | 81.1% | estimated ± 2.4 pp, medium confidence |
| Agents' Last Exam | 41.3% | estimated ± 6.3 pp, medium confidence |
| OpenHarmony Bench | 58.8% | estimated ± 2.4 pp, high confidence |
| MMLU | 94.6% | estimated ± 1.1 pp, low confidence |