Google releases Gemini 4 Carbon model with focus on cybersecurity
Google has launched Gemini 4 Carbon, a new AI model designed to improve coding and security capabilities, currently available to select defense partners.
Google announced the release of its latest artificial intelligence model, Gemini 4 Carbon, on Wednesday. The company reports performance improvements in coding, agentic workflows, and cybersecurity compared to its predecessor, Argon.
The model currently leads the Artificial Analysis Intelligence Index v5.0 benchmark, positioning it ahead of competitors such as Claude Opus 5.5 and GPT-6 Astra. While the index provides a quantitative metric for performance, the practical utility of these models in production environments remains a point of industry evaluation.
Access to Gemini 4 Carbon is currently restricted to a specific group of cybersecurity firms. Google plans to broaden availability through its Fairwind Program in future updates. This phased rollout reflects a strategy of limiting access to powerful frontier models to address potential security and safety concerns.