Models × Benchmarks
Compare
Build a head-to-head
Search any model to start, then add up to 4 and see them benchmark-by-benchmark. Or scan the overview matrix below.
| Model | GPQA Diamond | MMLU-Pro | IFEval | BIG-Bench Hard | MuSR | MATH Lvl 5 | Humanity's Last Exam | SciCode | AA Long-Context Reasoning | IFBench | τ²-bench | Terminal-Bench Hard |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 3.1 Pro Preview | 94.1 | · | · | · | · | · | 44.7 | 58.9 | 72.7 | 77.1 | 95.6 | 53.8 |
| GPT-5.6 Luna (max) | 91.1 | · | · | · | · | · | 37.2 | 52.5 | 74.0 | · | · | · |
| GPT-5.5 | 93.5 | · | · | · | · | · | 44.3 | 56.1 | 74.3 | 75.9 | 93.9 | 60.6 |
| Claude Sonnet 5 | 91.1 | · | · | · | · | · | 39.6 | 53.6 | 70.7 | · | · | · |
| Grok 4.5 | 93.1 | · | · | · | · | · | 40.3 | 54.1 | 67.7 | · | · | · |
| GLM-5.2 (max) | 89.5 | · | · | · | · | · | 40.1 | 50.5 | 71.3 | 73.3 | 99.1 | 50.8 |
| Qwen3.7 Max | 92.3 | · | · | · | · | · | 38.1 | 48.8 | 69.0 | 80.5 | 94.7 | 50.8 |
| GPT-5.4 | 92.0 | · | · | · | · | · | 41.6 | 56.6 | 74.0 | 73.9 | 87.1 | 57.6 |
| Claude Fable 5 | 92.6 | · | · | · | · | · | 53.3 | 60.2 | 70.0 | 63.5 | 98.5 | 62.9 |
| Kimi K2.6 | 91.1 | · | · | · | · | · | 35.9 | 53.5 | 69.7 | 76.0 | 95.9 | 43.9 |
| GPT-5.2 | 90.3 | 87.4 | · | · | · | · | 35.4 | 52.1 | 72.7 | 75.4 | 84.8 | 47.0 |
| Gemini 3.5 Flash | 92.2 | · | · | · | · | · | 41.0 | 53.1 | 69.3 | 76.3 | 95.3 | 40.9 |
| GPT-5.6 Sol (max) | 94.1 | · | · | · | · | · | 47.2 | 56.1 | 73.7 | 72.7 | 85.1 | 65.9 |
| GPT-5.6 Luna | 89.5 | · | · | · | · | · | 35.6 | 50.0 | 69.7 | · | · | · |
| MiMo-V2.5-Pro | 86.6 | · | · | · | · | · | 33.8 | 50.2 | 73.3 | 79.9 | 94.2 | 43.2 |
| GPT-5.2 Codex | 89.9 | · | · | · | · | · | 33.5 | 54.6 | 75.7 | 77.6 | 92.1 | 37.1 |
| Gemini 3 Pro Preview | 90.8 | 89.8 | · | · | · | · | 37.2 | 56.1 | 70.7 | 70.4 | 87.1 | 41.7 |
| Qwen3.6 Max Preview | 88.8 | · | · | · | · | · | 28.9 | 46.9 | 69.7 | 76.6 | 95.9 | 43.9 |
| GPT-5.3 Codex | 91.5 | · | · | · | · | · | 39.9 | 53.2 | 74.0 | 75.4 | 86.0 | 53.0 |
| Muse Spark 1.1 | 89.8 | · | · | · | · | · | 45.1 | 58.2 | 63.3 | · | · | · |
| DeepSeek V4 Pro | 88.8 | · | · | · | · | · | 35.9 | 50.0 | 66.3 | 76.5 | 96.2 | 46.2 |
| MiniMax-M3 | 92.9 | · | · | · | · | · | 37.1 | 45.4 | 74.0 | 82.9 | 88.9 | 42.4 |
| Muse Spark | 88.4 | · | · | · | · | · | 39.9 | 51.5 | 69.7 | 75.9 | 91.5 | 45.5 |
| GPT-5.6 Terra (max) | 92.5 | · | · | · | · | · | 41.8 | 53.9 | 74.0 | 71.2 | 86.3 | 57.6 |
| GPT-5.6 Sol | 93.1 | · | · | · | · | · | 44.7 | 56.0 | 71.0 | 71.0 | 84.8 | 61.4 |
| Gemini 3 Flash Preview | 89.8 | 89.0 | · | · | · | · | 34.7 | 50.6 | 66.3 | 78.0 | 80.4 | 38.6 |
| Claude Opus 4.8 | 92.0 | · | · | · | · | · | 45.7 | 53.5 | 67.7 | 62.2 | 94.4 | 58.3 |
| Grok 4.3 | 90.1 | · | · | · | · | · | 35.0 | 47.3 | 64.3 | 81.3 | 97.7 | 37.9 |
| Grok Build 0.1 0616 | 89.5 | · | · | · | · | · | 36.0 | 50.2 | 64.7 | · | · | · |
| GPT-5.1 | 87.3 | 87.0 | · | · | · | · | 26.5 | 43.3 | 75.0 | 72.9 | 81.9 | 45.5 |
· means not measured (distinct from a real 0). Click a model name to start a focused comparison.