| Benchmark | Model | Rank | Results |
|---|---|---|---|
| code-generation-on-deepswe | MiMo-V2.5 (CodeMidas RL) | – | Pass@1: 0.217 |
| code-generation-on-swe-bench-pro | MiMo-V2.5 (CodeMidas RL) | – | Resolved: 54.4 |
| code-generation-on-terminal-bench-2-1 | MiMo-V2.5 (CodeMidas RL) | – | Accuracy: 72.2 |