benchmark
Verified benchmarks
leaderboard
Hardcode-SWE (2026-11)
| model | bar | s |
|---|---|---|
| kimi-k3 | 91.3 | |
| deepseek-v4.1-flash | 89.2 | |
| glm-5.3 | 86.6 | |
| doubao-seed-2.1-lite | 85.0 | |
| glm-5.3-flash | 85.0 | |
| doubao-seed-2.1-pro | 83.0 | |
| kimi-k2.7-code | 80.5 | |
| mimo-v2.6-flash | 80.5 | |
| minimax-m3.1-flash | 78.5 | |
| glm-5.2 | 75.6 | |
| tencent-hy3 | 62.1 | |
| minimax-m2.7 | 59.9 | |
| tencent-hy4 | 57.9 |
Need a tailored sandbox, benchmark, or reinforcement learning environment? Email research@logitly.ai