reasoning benchmark
SWE-Bench Verified
A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.
Updated Aug 11, 2026
Models100
Model coverage105
MetricScore
EvidenceC
SWE-Bench Verified Ranking
Higher score ranks better on this benchmark.
SWE-Bench Verified Score Distribution
A closer view of the leading scores on this benchmark.
SWE-Bench Verified Highlights
The leading models and scores on this benchmark.
What is SWE-Bench Verified?
What SWE-Bench Verified measures and how its scores work.
A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of C.
- Family
- SWE-Bench Verified
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- swe-bench-verified|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about SWE-Bench Verified.
Which model scores highest on SWE-Bench Verified?
Claude Fable 5 is currently ranked first with 95.0%.
What does SWE-Bench Verified measure?
A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
100 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.