OpenAI · model
GPT-4 Turbo benchmark scores
As of 2026-10-07, GPT-4 Turbo (OpenAI) has measured scores on 3 benchmarks and estimated scores on 47 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 52.4% | estimated ± 6.8 pp, medium confidence |
| AA-HLE | 3.1% | measured |
| AA-SciCode | 32.1% | estimated ± 3.2 pp, high confidence |
| Artificial Analysis Intelligence Index | 7.0% | measured |
| CritPt | 0.3% | estimated ± 2.3 pp, high confidence |
| GPQA Diamond | 67.7% | estimated ± 2.2 pp, medium confidence |
| HLE | 19.4% | estimated ± 9.9 pp, low confidence |
| LiveCodeBench v6 | 70.8% | estimated ± 3.1 pp, high confidence |
| AA-Omniscience Accuracy | 17.3% | estimated ± 6.9 pp, medium confidence |
| SciCode | 32.4% | estimated ± 4.7 pp, high confidence |
| AA Coding Index | 21.5% | measured |
| BioMysteryBench (human-difficult) | 35.2% | estimated ± 8.1 pp, low confidence |
| SWE-bench Verified | 38.3% | estimated ± 5.5 pp, medium confidence |
| SWE-bench Pro | 28.6% | estimated ± 6.0 pp, low confidence |
| MMLU-Pro | 71.8% | estimated ± 4.5 pp, high confidence |
| GPQA | 44.5% | estimated ± 3.2 pp, medium confidence |
| ARC-AGI-1 | 0.4% | estimated ± 5.4 pp, low confidence |
| ARC-AGI-2 | 0.2% | estimated ± 14.2 pp, low confidence |
| CursorBench 3.1 | 47.9% | estimated ± 4.8 pp, low confidence |
| CursorBench 3.2 | 12.4% | estimated ± 4.4 pp, medium confidence |
| FrontierCode 1.1 Main | 6.6% | estimated ± 3.8 pp, medium confidence |
| FrontierSWE v2 | 0.0% | estimated ± 14.8 pp, low confidence |
| Vals GPQA Diamond | 3.0% | estimated ± 7.5 pp, low confidence |
| Vals MMLU-Pro | 83.1% | estimated ± 2.5 pp, medium confidence |
| Vals SWE-bench | 49.2% | estimated ± 5.4 pp, low confidence |
| CursorBench 4.0 | 4.8% | estimated ± 1.6 pp, medium confidence |
| DeepSWE | 0.1% | estimated ± 6.4 pp, low confidence |
| HLE w/o tools | 2.4% | estimated ± 5.7 pp, low confidence |
| PostTrainBench v1.1 | 8.3% | estimated ± 6.1 pp, low confidence |
| Vibe Code Bench | 6.1% | estimated ± 10.3 pp, low confidence |
| C-Eval | 77.9% | estimated ± 1.0 pp, low confidence |
| MMLU-Redux | 71.2% | estimated ± 2.1 pp, low confidence |
| NL2Repo | 13.9% | estimated ± 5.1 pp, low confidence |
| SuperGPQA | 34.4% | estimated ± 10.4 pp, low confidence |
| HealthBench Hard | 0.0% | estimated ± 7.7 pp, low confidence |
| MMLU-Pro (Arcee) | 77.3% | estimated ± 5.4 pp, low confidence |
| React Native Evals | 69.5% | estimated ± 3.5 pp, high confidence |
| SWE-Rebench | 0.0% | estimated ± 8.9 pp, low confidence |
| ARC-AGI-3 | 1.6% | estimated ± 10.7 pp, low confidence |
| BioMysteryBench (human-solvable) | 0.1% | estimated ± 1.0 pp, low confidence |
| FrontierCode 1.1 Extended | 21.6% | estimated ± 3.4 pp, low confidence |
| HealthBench (raw) | 4.9% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 3.7 pp, low confidence |
| HealthBench Professional | 57.7% | estimated ± 4.8 pp, low confidence |
| HealthBench Professional (raw) | 19.4% | estimated ± 5.0 pp, low confidence |
| HLE-Verified | 0.0% | estimated ± 6.8 pp, low confidence |
| LABBench2 | 64.1% | estimated ± 2.4 pp, low confidence |
| LiveCodeBench | 35.0% | estimated ± 8.9 pp, low confidence |
| OpenHarmony Bench | 34.6% | estimated ± 2.8 pp, medium confidence |
| MMLU | 78.8% | estimated ± 1.9 pp, low confidence |