Benchmark

CW Bench

A benchmark focused on practical cybersecurity capability that requires models to locate the root cause of vulnerabilities and provide remediation plans.

Core Facts

Gemini 4 Carbon achieved a score of 71% on the CW Bench cybersecurity test.

Partially Verified65% confidence3 sources

Timeline (last 90 days)

Oct 29

Gemini 4 Carbon achieved a score of 71% on the CW Bench cybersecurity test.

Partially Verified65%

Source Articles

Google Releases Gemini 4 Carbon: Security Agent Back on TopOct 29 Gemini 4 Carbon Hands-On: The Strongest AI Model? Compared with Claude Opus 5.5Oct 29