| Rank | Model | Provider | Score |
|---|---|---|---|
| 1 | GPT 6 Astra | OpenAI | 85.4% |
| 2 | GPT 5.6 Sol | OpenAI | 74.3% |
| 3 | MiMo V2.6 Pro | Xiaomi | 66.3% |
| 4 | DeepSeek V4.1 Flash | DeepSeek | 62.8% |
| 5 | DeepSeek V4 Pro 0813 | DeepSeek | 56.4% |
| 6 | MiMo V2.6 Flash | Xiaomi | 47.5% |
| 7 | DeepSeek V4 Flash 0731 | DeepSeek | 30.9% |
| 8 | MiMo V2.5 Pro | Xiaomi | 17.7% |
Figures are taken from public model cards, technical reports and vendor announcements, and are approximate where sources disagree. Results from different benchmark versions or tool-use settings never share a column: Terminal Bench 2.1, 3.0 and DeepSWE v1.1 are distinct benchmarks, and HLE with tools is separate from HLE without tools. Only models with a published result are listed; the score column is shaded relative to the best and worst result on this benchmark.