Benchmark
SWE-bench Multilingual
A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632…
- Modality
- text
- Categories
- reasoning, code
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 38
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Claude Mythos Preview | Anthropic | 0.873 | 2026-04-07 | evidence |
| Claude Opus 4.6 | Anthropic | 0.7783 | 2026-02-05 | evidence |
| Claude Opus 4.8 | Anthropic | 0.844 | 2026-05-28 | evidence |
| Claude Sonnet 5 | Anthropic | 0.783 | 2026-06-30 | evidence |
| DeepSeek-R1-0528 | DeepSeek | 0.305 | 2025-05-28 | evidence |
| DeepSeek-V3.1 | DeepSeek | 0.545 | 2025-01-10 | evidence |
| DeepSeek-V3.2 | DeepSeek | 0.702 | 2025-12-01 | evidence |
| DeepSeek-V3.2 (Thinking) | DeepSeek | 0.702 | 2025-12-01 | evidence |
| DeepSeek-V3.2-Exp | DeepSeek | 0.579 | 2025-09-29 | evidence |
| DeepSeek-V4-Flash-0423 | DeepSeek | 0.702 | 2026-04-23 | evidence |
| DeepSeek-V4-Flash-Max | DeepSeek | 0.733 | 2026-04-23 | evidence |
| DeepSeek-V4-Pro-Max | DeepSeek | 0.762 | 2026-04-23 | evidence |
| GLM-4.7 | Z.ai | 0.667 | 2025-12-22 | evidence |
| Hy3 | Tencent | 0.758 | 2026-07-06 | evidence |
| Kimi K2 Instruct | Moonshot AI | 0.473 | 2025-07-11 | evidence |
| Kimi K2-Instruct-0905 | Moonshot AI | 0.473 | 2025-09-05 | evidence |
| Kimi K2-Thinking-0905 | Moonshot AI | 0.611 | 2025-09-05 | evidence |
| Kimi K2.5 | Moonshot AI | 0.73 | 2026-01-27 | evidence |
| Kimi K2.6 | Moonshot AI | 0.767 | 2026-04-20 | evidence |
| Laguna S 2.1 | Poolside | 0.785 | 2026-07-21 | evidence |
| Laguna XS 2.1 | Poolside | 0.631 | 2026-07-02 | evidence |
| LongCat-Flash-Lite | Meituan | 0.381 | 2026-02-05 | evidence |
| MAI-Code-1-Flash | Microsoft | 0.655 | 2026-06-02 | evidence |
| MiMo-V2-Flash | Xiaomi | 0.717 | 2025-12-16 | evidence |
| MiMo-V2-Pro | Xiaomi | 0.717 | 2026-03-18 | evidence |
| MiniMax M2 | MiniMax | 0.565 | 2025-10-27 | evidence |
| MiniMax M2.1 | MiniMax | 0.725 | 2025-12-23 | evidence |
| MiniMax M2.7 | MiniMax | 0.765 | 2026-03-18 | evidence |
| Nemotron 3 Super (120B A12B) | NVIDIA | 0.4578 | 2026-03-11 | evidence |
| Nemotron 3 Ultra (550B A55B) | NVIDIA | 0.677 | 2026-06-04 | evidence |
| Nemotron 3.5 Lightning (30B A3B) | NVIDIA | 0.3933 | 2026-08-11 | evidence |
| Qwen3-Coder 480B A35B Instruct | Qwen | 0.547 | 2025-01-31 | evidence |
| Qwen3.5-397B-A17B | Qwen | 0.693 | 2026-02-16 | evidence |
| Qwen3.6 Plus | Qwen | 0.738 | 2026-04-02 | evidence |
| Qwen3.6-27B | Qwen | 0.713 | 2026-04-21 | evidence |
| Qwen3.6-35B-A3B | Qwen | 0.672 | 2026-04-16 | evidence |
| Qwen3.7 Max | Qwen | 0.783 | 2026-05-19 | evidence |
| Qwen3.7-Plus | Qwen | 0.758 | 2026-05-31 | 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.