MAI-Thinking-1 ranks first on 2 of the 20 benchmarks with five or more models scored.
Microsoft Jun 2026 Flagship
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| TruthfulQA | 88% | #1 of 30 | Lab model card | |
| AdvancedIF | 85% | #1 of 6 | Lab model card | |
| MMLU-Pro | 85% | #31 of 167 | Lab model card | |
| GraphWalks | 90% | #7 of 27 | Lab model card | |
| LongBench v2 | 61% | #12 of 35 | Lab model card | |
| IFBench | 69% | #47 of 80 | Lab model card | |
| OpenAI MRCR v2 | 60% | #9 of 14 | Lab model card | |
| MultiChallenge | 53% | #29 of 42 | Lab model card | |
| SimpleQA Verified | 31% | #18 of 22 | Lab model card | |
| LongFact | 98% | #5 of 5 | Lab model card | |
| CorpusQA | 82% | #1 of 1 | Lab model card | |
| MAI human side-by-side evaluation | 0.07 | #1 of 1 | Lab model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| AIME 2025 | 97% | #17 of 143 | Lab model card | |
| GPQA Diamond | 84.2% | #85 of 302 | Lab model card | |
| AIME 2026 | 94.5% | #15 of 38 | Lab model card | |
| HMMT 2026 | 84.9% | #13 of 19 | Lab model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LiveCodeBench v6 | 87.7% | #12 of 90 | Lab model card | |
| SWE-bench Verified | 73.5% | #55 of 155 | Lab model card | |
| SWE-Bench Pro | 52.8% | #59 of 81 | Lab model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Berkeley Function Calling Leaderboard (BFCL) | 72% | #10 of 41 | Lab model card | |
| Terminal-Bench 2.0 | 46% | #45 of 63 | Lab model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HealthBench Professional | 35% | #26 of 26 | Lab model card | |
| MedXpertQA | 43% | #2 of 2 | Lab model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| AIR-Bench 2024 | 88% | #1 of 2 | Lab model card | |
| CyberSecEval Autocomplete | 63% | #2 of 3 | Lab model card | |
| CyberSecEval Instruct | 63% | #3 of 3 | Lab model card |
The Self-Reported Index fits one score to the 18 capability results Microsoft reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 142.6 to 147.3. How we compute it.
The benchmarks the lab chose to headline, with the scores it gave.
Every result in the longer documents the lab published with it. 9 of them aren't in the launch post.