Martian Unbiased OSS Benchmark

The #1 Ranked
AI Code Reviewer

Beats Copilot, Claude, Devin, CodeRabbit and 12 others on real PRs.

START FREE TRIAL

NO CC REQUIRED

See the bechmark

Trusted by Startups to Fortune 500

Martian Independent Benchmark

Full Leaderboard

Security Patch Detection — F1 Score (%)

`#1 of 20 tools`

706050403020100

66.7%

CodeAnt AI

66.7%

Qodo Extended

57.1%

Claude

54.5%

Cubic Dev

50%

Devin

706050403020100

66.7%

57.1%

54.5%

50%

706050403020100

66.7%

57.1%

54.5%

50%

Logging & PII Leak Detection — F1 Score (%)

`#1 of 20 tools`

80706050403020100

71.4%

CodeAnt AI

71.4%

Qodo Extended

58.8%

Code Rabbit

55.6%

Augment

50%

Cursor

80706050403020100

71.4%

58.8%

55.6%

50%

80706050403020100

71.4%

58.8%

55.6%

50%

Large & Medium PR Reviews — F1 Score (%)

`#1 of 20 tools`

706050403020100

67%

Qodo Extended

55%

CodeAnt AI

50.6%

Augment

45.4%

Cursor

45.1%

Qodo

706050403020100

67%

55%

50.6%

45.4%

45.1%

706050403020100

67%

55%

50.6%

45.4%

45.1%

Missing Test Coverage — F1 Score (%)

`#1 of 20 tools`

706050403020100

66.7%

CodeAnt AI

40%

Greptile

33.3%

Cubic Dev

Augment

Claude

706050403020100

66.7%

40%

33.3%

706050403020100

66.7%

40%

33.3%

Source - Martian's Independent Benchmark

Leaderboard

#	Tool	F1	Precision	Recall
1	CodeAnt AI	66.7%	66.7%	66.7%
2	Qodo Extended	66.7%	66.7%	66.7%
3	Claude Code Reviewer	57.1%	50.0%	66.7%
4	Cubic Dev	54.5%	37.5%	100%
5	Devin	50.0%	100%	33.3%
6	Qodo	50.0%	40.0%	66.7%
7	Baz	40.0%	50.0%	33.3%
8	Claude Code Plugin	40.0%	50.0%	33.3%
9	Gemini	40.0%	50.0%	33.3%
10	Augment	33.3%	33.3%	33.3%