Unveiled Gemini 4 Carbon, a frontier model focused on cyber defense. (Oct. 28)

On Oct. 28, Google unveiled Gemini 4 Carbon, its latest frontier model, positioning it around cybersecurity defense and long-running coding work. Google says Carbon ranks first on Artificial Analysis Intelligence Index v5.0, scoring 59, and reports 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for testing exploitation, and 71.0% on Collinear’s CWE-bench for producing patches. The company also says Carbon generates secure code at more than twice Gemini 4 Argon’s success rate on internal tests. These claims put Google directly against OpenAI and Anthropic in the fast-growing market for AI-assisted software security; defenders could use the model to identify, validate and repair flaws, while its offensive capabilities make controlled access consequential. Google is initially rolling Carbon out to a subset of cyber defenders through its Fairwind Program, with broader availability planned but no public date stated.