reasoning benchmark
Multi-SWE-Bench
A multilingual benchmark for issue resolving that evaluates Large Language Models' ability to resolve software issues across diverse programming ecosystems. Covers 7 programming languages (Java, TypeScript, JavaScript, Go, Rust, C, and C++) with 1,632 high-quality instances carefully annotated by 68 expert annotators. Addresses limitations of existing benchmarks that focus almost exclusively on Python.
Updated Aug 7, 2026
Published models6
Registry coverage6
MetricScore
EvidenceB
Multi-SWE-Bench leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MI | 52.7% | 100.0% | 6 | C | |
| 2 | MI | 51.3% | 80.0% | 6 | C | |
| 3 | MI | 49.4% | 60.0% | 6 | C | |
| 4 | MA | 41.9% | 40.0% | 6 | C | |
| 5 | MI | 36.2% | 20.0% | 6 | C | |
| 6 | AC | 25.8% | 0.0% | 6 | C |
Score distribution
Top published rows on the benchmark's original scale.
Multi-SWE-Bench highlights
The top published results on this benchmark's own scale.
What is Multi-SWE-Bench?
Definition and scoring fields from the benchmark registry.
A multilingual benchmark for issue resolving that evaluates Large Language Models' ability to resolve software issues across diverse programming ecosystems. Covers 7 programming languages (Java, TypeScript, JavaScript, Go, Rust, C, and C++) with 1,632 high-quality instances carefully annotated by 68 expert annotators. Addresses limitations of existing benchmarks that focus almost exclusively on Python.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- Multi-SWE-Bench
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- multi-swe-bench|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about Multi-SWE-Bench.
Which model scores highest on Multi-SWE-Bench?
MiniMax M2.7 is currently ranked first with 52.7%.
What does Multi-SWE-Bench measure?
A multilingual benchmark for issue resolving that evaluates Large Language Models' ability to resolve software issues across diverse programming ecosystems. Covers 7 programming languages (Java, TypeScript, JavaScript, Go, Rust, C, and C++) with 1,632 high-quality instances carefully annotated by 68 expert annotators. Addresses limitations of existing benchmarks that focus almost exclusively on Python.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
6 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.