Benchmark
Multi-SWE-Bench
A multilingual benchmark for issue resolving that evaluates Large Language Models' ability to resolve software issues across diverse programming…
- Modality
- text
- Categories
- reasoning, code
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 6
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Kimi K2-Thinking-0905 | Moonshot AI | 0.419 | 2025-09-05 | evidence |
| MiniMax M2 | MiniMax | 0.362 | 2025-10-27 | evidence |
| MiniMax M2.1 | MiniMax | 0.494 | 2025-12-23 | evidence |
| MiniMax M2.5 | MiniMax | 0.513 | 2026-02-12 | evidence |
| MiniMax M2.7 | MiniMax | 0.527 | 2026-03-18 | evidence |
| Qwen3-Coder 480B A35B Instruct | Qwen | 0.258 | 2025-01-31 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.