Code Review Assistant
Reviewing pull requests, identifying bugs, security issues, and style violations.
Best for this use case
claude-sonnet-4
Strong on Sonar Java Quality Leaderboard functional_skill_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct
23.2%
Best benchmark score
32.6%
Confidence
All ranked models — top 3
Ranked Models
30
Evidence Quality
90%
Evidence Points
29
Top Signal
Sonar Java Quality Leaderboard: functional_skill_pct
Benchmark Sources
45
Last Updated
6h ago
Benchmark Sources (7)
All Ranked Models
| Rank | Model | Score |
|---|---|---|
| 🥇 | claude-sonnet-4 Strong on Sonar Java Quality Leaderboard functional_skill_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct | 23.2% |
| 🥈 | gpt-5-2025-08-07 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and SWE-bench Leaderboard verified_resolved_pct | 21.9% |
| 🥉 | gemini-3-pro-preview Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and SWE-bench Leaderboard verified_resolved_pct | 21.8% |
| #4 | Kimi K2 Thinking Strong on Sonar Java Quality Leaderboard functional_skill_pct and SWE-bench Leaderboard verified_resolved_pct | 19.1% |
| #5 | o3-20250416 Strong on SWE-bench Leaderboard verified_resolved_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct | 18.2% |
| #6 | gpt-5.2-2025-12-11 Strong on SWE-bench Leaderboard verified_resolved_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct | 18.1% |
| #9 | gemini-2.5-pro Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and SWE-bench Leaderboard verified_resolved_pct | 17.1% |
| #11 | claude-opus-4-5-20251101 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and SWE-bench Leaderboard verified_resolved_pct | 16.3% |
| #12 | deepseek-r1 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc | 16.2% |
| #13 | gpt-4o Strong on TestEval Leaderboard overall_average_coverage_pct and Sonar Java Quality Leaderboard functional_skill_pct | 16.0% |
| #14 | gpt-4.1-20250414 Strong on Galileo Agent Leaderboard v2 Avg AC and SWE-bench Leaderboard verified_resolved_pct | 15.7% |
| #15 | qwen-2.5-coder32b-instruct Strong on BigCode Models Leaderboard average_score and BigCodeBench Official bigcodebench_complete_pct | 15.2% |
| #16 | o4-mini Strong on SWE-bench Leaderboard verified_resolved_pct and Berkeley Function Calling Leaderboard (Overall) Overall Acc | 15.2% |
| #17 | gpt-5-mini-2025-08-07 Strong on SWE-bench Leaderboard verified_resolved_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct | 15.0% |
| #18 | GLM-5 Strong on Sonar Java Quality Leaderboard functional_skill_pct and SWE-bench Leaderboard verified_resolved_pct | 14.8% |
| #19 | claude-opus-4-6-thinking Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc | 14.6% |
| #20 | qwen-2.5-72b-instruct Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and BigCodeBench Official bigcodebench_complete_pct | 14.0% |
| #25 | GLM-4.6 Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and SWE-bench Verified Leaderboard swe_verified_resolved_pct | 12.9% |
| #28 | Kimi-K2-Instruct Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and SWE-bench Leaderboard verified_resolved_pct | 12.5% |
| #29 | gemini-3-flash-preview Strong on SWE-bench Leaderboard verified_resolved_pct and Vals LiveCodeBench overall_accuracy_pct | 12.3% |
| #30 | gemini-2.5-flash Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and OSS-Bench Leaderboard average_score_pct | 12.2% |
| #32 | Phi-3-medium-128k-instruct Strong on RepoQA Official Results overall_average_pass_at_1_pct and Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct | 11.7% |
| #33 | Grok-4-0709 Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and Vals LiveCodeBench overall_accuracy_pct | 11.7% |
| #34 | glm-4.7 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc | 11.5% |
| #40 | minimax-m2.1 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc | 10.7% |
| #41 | gpt-5.1-2025-11-13 Strong on SWE-bench Leaderboard verified_resolved_pct and Vals LiveCodeBench overall_accuracy_pct | 10.7% |
| #42 | Mixtral-8x22B-Instruct-v0.1 Strong on RepoQA Official Results overall_average_pass_at_1_pct and Open LLM Leaderboard GPQA gpqa | 10.5% |
| #44 | kimi-k2.5-thinking Strong on SWE-bench Leaderboard verified_resolved_pct and Vals LiveCodeBench overall_accuracy_pct | 10.4% |
| #51 | Qwen/Qwen1.5-32B-Chat Strong on RepoQA Official Results overall_average_pass_at_1_pct and Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct | 9.5% |
| #52 | grok-4-1-fast-reasoning Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and Vals LiveCodeBench overall_accuracy_pct | 9.4% |
Compare Models
▶Ranking diagnostics & missing models
Source lift
Ranked
84
Sources
8
Quality
Good
Open LLM Leaderboard MMLU-Pro
Open LLM Leaderboard GPQA
Open LLM Leaderboard BBH
Open LLM Leaderboard Results
Missing frontier models
claude-sonnet-4.6
Thin evidence after weightingRank #11
17.9%
gemini-3.1-pro-preview
Thin evidence after weightingRank #17
20.8%
grok-4-fast-reasoning
Thin evidence after weightingRank #19
14.0%
▶Taxonomy & task details
Core tasks
Required modes
Domains
Related in Developer
Autonomous Coding Agent
End-to-end autonomous software engineering: reading issues, writing code, running tests, submitting PRs.
Function Calling / Tool Use Agent
Reliably selecting and invoking the correct tools/APIs in multi-step agentic pipelines.
PR review agent
Review diffs for correctness, security, and maintainability.
Agentic bug fixing
Agentic loop that reproduces, fixes, and validates with tests.