reasoning benchmark
SWE-bench Verified (Agentic Coding) Leaderboard
SWE-bench Verified is a human-filtered subset of 500 software engineering problems drawn from real GitHub issues across 12 popular Python repositories. Given a codebase and an issue description, language models are tasked with generating patches that resolve the described problems. This benchmark evaluates AI's real-world agentic coding skills by requiring models to navigate complex codebases, understand software engineering problems, and coordinate changes across multiple functions, classes, and files to fix well-defined issues with clear descriptions.
Updated Aug 17, 2026
Models2
Model coverage2
MetricScore
EvidenceB
SWE-bench Verified (Agentic Coding) Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAN | Score77.2% | Percentile100.0% | Participants2 | EvidenceC | Evaluated |
| Rank02 | ModelMA | Score65.8% | Percentile0.0% | Participants2 | EvidenceC | Evaluated |
SWE-bench Verified (Agentic Coding) Highlights
The leading models and scores on this benchmark.
SWE-bench Verified (Agentic Coding) Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for SWE-bench Verified (Agentic Coding)
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis swe-bench verified (agentic coding) AI model leaderboard uses descending score in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01ANClaude Sonnet 4.5Anthropic
- Score
- 77.2%
- Price
- $3.0 input / $15 output per 1M tokens
- Speed
- Up to 42 tok/s via Anthropic
Strengths
- Ranks #1 of 2 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures SWE-bench Verified (Agentic Coding), not total model capability
- 02MAKimi K2 InstructMoonshot AI
- Score
- 65.8%
- Price
- $0.60 input / $2.5 output per 1M tokens
- Speed
- Up to 45 tok/s via Novita
Strengths
- Ranks #2 of 2 compared models
- 0th percentile on this benchmark
- C evidence result
Considerations
- This result measures SWE-bench Verified (Agentic Coding), not total model capability
Selection summary
Best AI Models for SWE-bench Verified (Agentic Coding)
Claude Sonnet 4.5 currently leads SWE-bench Verified (Agentic Coding) with 77.2%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is SWE-bench Verified (Agentic Coding)?
What SWE-bench Verified (Agentic Coding) measures and how its scores work.
SWE-bench Verified is a human-filtered subset of 500 software engineering problems drawn from real GitHub issues across 12 popular Python repositories. Given a codebase and an issue description, language models are tasked with generating patches that resolve the described problems. This benchmark evaluates AI's real-world agentic coding skills by requiring models to navigate complex codebases, understand software engineering problems, and coordinate changes across multiple functions, classes, and files to fix well-defined issues with clear descriptions.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- SWE-bench Verified (Agentic Coding)
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- swe-bench-verified-(agentic-coding)|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about SWE-bench Verified (Agentic Coding).
Which model scores highest on SWE-bench Verified (Agentic Coding)?
Claude Sonnet 4.5 is currently ranked first with 77.2%.
What does SWE-bench Verified (Agentic Coding) measure?
SWE-bench Verified is a human-filtered subset of 500 software engineering problems drawn from real GitHub issues across 12 popular Python repositories. Given a codebase and an issue description, language models are tasked with generating patches that resolve the described problems. This benchmark evaluates AI's real-world agentic coding skills by requiring models to navigate complex codebases, understand software engineering problems, and coordinate changes across multiple functions, classes, and files to fix well-defined issues with clear descriptions.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.