x.comthread🚨 AI News | TestingCatalog (@testingcatalog) on X
Gemini 4 Carbon is a new Google model, announced in a TestingCatalog post on X as taking first place on the Artificial Analysis V5 ranking. It is rolling out to cyber defenders first. Reported scores are 88.2% on SEC-bench Pro (finding vulnerabilities), 54.8% on ExploitGym (exploitation) and 71.0% on CWE-bench (patching). The post calls it a large step up from Gemini 4 Argon, released two weeks earlier. It gives no technical detail beyond these claims.
TL;DR Announces Google's Gemini 4 Carbon, ranked first on Artificial Analysis V5 with strong cyber benchmark scores and a staged rollout to defenders.
Organisations named here
- Artificial Analysis
- TestingCatalog
- CWE-bench
- SEC-bench Pro
Technologies named here
- Gemini 4 Carbon
- Gemini 4 Argon
- Gemini
- ExploitGym
Related in models
