Developer

Code Review Assistant

Reviewing pull requests, identifying bugs, security issues, and style violations.

task.code_reviewtask.risk_assessment

Evidence quality is currently limited for this use case. Rankings below are useful for exploration, not a strong winner claim.

Provisional leader

claude-sonnet-4

Current leader based on limited benchmark evidence. Treat this ranking as directional until coverage improves.

27.1%

Best benchmark score

38.8%

Confidence

All ranked models — top 3

🥇

claude-sonnet-4

27.1%

🥈

gpt-5-2025-08-07

24.8%

🥉

deepseek-r1

22.8%

Ranked Models

Evidence Quality

80%

Evidence Points

Top Signal

Sonar Java Quality Leaderboard: functional_skill_pct

All Ranked Models

30 of 30 models

Rank	Model	Score	Confidence	Price / 1M	Evidence sources
🥇	claude-sonnet-4 Strong on Sonar Java Quality Leaderboard functional_skill_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct	27.1%	39%	$6.00	Sonar Java Quality LeaderboardSWE-bench Verified Leaderboard
🥈	gpt-5-2025-08-07 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Aider Polyglot Leaderboard percent_correct_pct	24.8%	30%	—	SWE-bench Verified LeaderboardAider Polyglot Leaderboard
🥉	deepseek-r1 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc	22.8%	34%	$0.27	Sonar Java Quality LeaderboardSonar Java Quality Leaderboard
#4	gpt-4o-2024-05-13 Strong on RepoQA Official Results overall_average_pass_at_1_pct and BigCodeBench Official bigcodebench_complete_pct	22.6%	29%	—	RepoQA Official ResultsBigCodeBench Official
#7	Kimi K2 Thinking Strong on Sonar Java Quality Leaderboard functional_skill_pct and SWE-bench Verified Leaderboard swe_verified_resolved_pct	21.2%	30%	$1.07	Sonar Java Quality LeaderboardSWE-bench Verified Leaderboard
#8	gemini-2.5-pro Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and LEXam Leaderboard average_score_pct	19.3%	28%	$3.44	SWE-bench Verified LeaderboardLEXam Leaderboard
#9	gemini-3-pro-preview Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Vals SWE-bench overall_accuracy_pct	18.6%	23%	$4.50	SWE-bench Verified LeaderboardVals SWE-bench
#11	gpt-4o Strong on Sonar Java Quality Leaderboard functional_skill_pct and Aider Refactoring Leaderboard percent_correct_pct	16.7%	35%	$0.26	Sonar Java Quality LeaderboardAider Refactoring Leaderboard
#12	glm-4.7 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc	16.4%	24%	—	Sonar Java Quality LeaderboardSonar Java Quality Leaderboard
#13	o3-20250416 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Aider Polyglot Leaderboard percent_correct_pct	16.2%	22%	$3.50	SWE-bench Verified LeaderboardAider Polyglot Leaderboard
#14	gpt-5-mini-2025-08-07 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Vals LiveCodeBench overall_accuracy_pct	16.2%	23%	—	SWE-bench Verified LeaderboardVals LiveCodeBench
#16	gpt-4o-20241120 Strong on BigCodeBench Official bigcodebench_complete_pct and BigCodeBench Official bigcodebench_hard_complete_pct	15.8%	26%	—	BigCodeBench OfficialBigCodeBench Official
#17	gpt-5.2-2025-12-11 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Vals SWE-bench overall_accuracy_pct	15.6%	18%	—	SWE-bench Verified LeaderboardVals SWE-bench
#18	minimax-m2.1 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Vals SWE-bench overall_accuracy_pct	15.3%	24%	$0.53	Sonar Java Quality LeaderboardVals SWE-bench
#19	gpt-4.1-20250414 Strong on Galileo Agent Leaderboard v2 Avg AC and SWE-bench Verified Leaderboard swe_verified_resolved_pct	15.2%	24%	—	Galileo Agent Leaderboard v2SWE-bench Verified Leaderboard
#20	gpt-4.1 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Aider Polyglot Leaderboard percent_correct_pct	15.1%	19%	$3.50	SWE-bench Verified LeaderboardAider Polyglot Leaderboard
#23	GLM-5 Strong on Sonar Java Quality Leaderboard functional_skill_pct and Sonar Java Quality Leaderboard issue_density_error_per_kloc	14.0%	15%	—	Sonar Java Quality LeaderboardSonar Java Quality Leaderboard
#27	claude-opus-4-5-20251101 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Vals Terminal-Bench 2 overall_accuracy_pct	13.1%	16%	—	SWE-bench Verified LeaderboardVals Terminal-Bench 2
#28	qwen-2.5-72b-instruct Strong on BigCodeBench Official bigcodebench_complete_pct and Galileo Agent Leaderboard v2 Avg AC	13.1%	18%	—	BigCodeBench OfficialGalileo Agent Leaderboard v2
#32	gemini-3.1-pro-preview Strong on Vals SWE-bench overall_accuracy_pct and Vals LiveCodeBench overall_accuracy_pct	12.0%	13%	$4.50	Vals SWE-benchVals LiveCodeBench
#34	claude-opus-4 Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Aider Polyglot Leaderboard percent_correct_pct	11.7%	15%	$10.00	SWE-bench Verified LeaderboardAider Polyglot Leaderboard
#36	o4-mini Strong on SWE-bench Verified Leaderboard swe_verified_resolved_pct and Aider Polyglot Leaderboard percent_correct_pct	11.6%	18%	$1.93	SWE-bench Verified LeaderboardAider Polyglot Leaderboard
#37	Phi-3-medium-128k-instruct Strong on RepoQA Official Results overall_average_pass_at_1_pct and BigCodeBench Official bigcodebench_complete_pct	11.5%	17%	—	RepoQA Official ResultsBigCodeBench Official
#43	Grok-4-0709 Strong on Vals LiveCodeBench overall_accuracy_pct and Galileo Agent Leaderboard v2 Avg AC	11.2%	18%	—	Vals LiveCodeBenchGalileo Agent Leaderboard v2
#47	deepseek-v3 Strong on BigCodeBench Official bigcodebench_complete_pct and BigCodeBench Official bigcodebench_hard_complete_pct	10.9%	16%	—	BigCodeBench OfficialBigCodeBench Official
#62	gemini-3-flash-preview Strong on Vals SWE-bench overall_accuracy_pct and Vals LiveCodeBench overall_accuracy_pct	10.1%	12%	$1.13	Vals SWE-benchVals LiveCodeBench
#64	gpt-4o-2024-08-06 Strong on BigCodeBench Official bigcodebench_hard_complete_pct and Aider Code Editing Leaderboard percent_correct_pct	10.1%	17%	—	BigCodeBench OfficialAider Code Editing Leaderboard
#65	gpt-5.4-2026-03-05 Strong on Vals SWE-bench overall_accuracy_pct and Vals LiveCodeBench overall_accuracy_pct	10.1%	11%	—	Vals SWE-benchVals LiveCodeBench
#68	claude-sonnet-4.6 Strong on Vals SWE-bench overall_accuracy_pct and Vals LiveCodeBench overall_accuracy_pct	9.8%	11%	$6.00	Vals SWE-benchVals LiveCodeBench
#70	claude-opus-4-6-thinking Strong on Vals SWE-bench overall_accuracy_pct and Vals LiveCodeBench overall_accuracy_pct	9.6%	10%	—	Vals SWE-benchVals LiveCodeBench

Compare Models

Select two different models above to compare their evidence side by side.

▶Ranking diagnostics & missing models

Source lift

Ranked

Sources

Quality

Low

Vals LiveCodeBench

35 rows · 0.9% avg lift

Vals Legal Bench

32 rows · 0.2% avg lift

Vals MedQA

31 rows · 0.2% avg lift

Vals Tax Eval v2

31 rows · 0.2% avg lift

Missing frontier models

grok-4-1-fast-non-reasoning

Thin evidence after weighting

Rank #15

14.9%

▶Taxonomy & task details

Core tasks

task.code_reviewtask.risk_assessment

Required modes

none

Domains

domain.software_engineering

Related in Developer

Codebase onboarding brief

Summarize a repository's architecture, modules, and conventions.

Autonomous Coding Agent

End-to-end autonomous software engineering: reading issues, writing code, running tests, submitting PRs.

Code generation

Generate correct, secure code from requirements.

Refactoring assistant

Refactor code safely while preserving behavior and improving clarity.