Google Unveils Gemini 4 Carbon, Its Most Powerful AI Model Yet

Google DeepMind launched Gemini 4 Carbon on October 28, 2026, describing it as the company’s strongest AI model to date. The model is designed with a particular emphasis on cybersecurity and long-horizon agentic tasks.

Carbon debuted at the top of the Artificial Analysis Intelligence Index v5.0, scoring 59 and outperforming all current rivals. Its cybersecurity benchmark results include:

• 88.2% on SEC-bench Pro for vulnerability discovery
• 54.8% on ExploitGym for exploitation
• 71.0% on CWE-bench for security patching

Rather than launching broadly to consumers, Google is taking a deliberate, phased approach. Initial access this week is limited to a select group of vetted cyber defenders through its Fairwind Program, which includes more than 650 trusted partners worldwide.

Koray Kavukcuoglu, SVP at Google DeepMind and Chief AI Architect, said the company is participating in the U.S. government’s voluntary pre-release model testing process. Broader access for developers, businesses, and consumers has been promised, but no firm date has been set.

Carbon arrives just two weeks after Gemini 4 Argon, underscoring Google’s aggressive pace of frontier AI development.

Source: Google Blog — Gemini 4 Carbon: Our Next Frontier Model