| code-generation-on-livecodebench-v6 | MAI-Thinking-1 | – | Pass@1: 87.7 |
| code-generation-on-swe-bench-pro | MAI-Thinking-1 | – | Resolved: 52.8 |
| code-generation-on-swe-bench-verified | MAI-Thinking-1 | – | Accuracy: 73.5 |
| instruction-following-on-ifbench | MAI-Thinking-1 | – | Accuracy: 69 |
| instruction-following-on-multichallenge | MAI-Thinking-1 | – | Accuracy: 53 |
| long-context-understanding-on-longbench-v2 | MAI-Thinking-1 | – | Accuracy: 61 |
| mathematical-reasoning-on-aime-2025 | MAI-Thinking-1 | – | Accuracy: 97.0 |
| mathematical-reasoning-on-aime-2026 | MAI-Thinking-1 | – | Accuracy: 94.5 |
| mathematical-reasoning-on-hmmt-feb-2026 | MAI-Thinking-1 | – | Accuracy: 84.9 |
| multi-task-language-understanding-on-mmlu-pro | MAI-Thinking-1 | – | Accuracy: 85 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | MAI-Thinking-1 | – | Accuracy: 84.2 |