Google makes Gemini 4 Carbon available to developers

What's new

Google announced on Nov 10, 2026 that developers can now build with Gemini 4 Carbon. The company calls it its strongest model and claims top-tier results in cybersecurity as well as in tasks that unfold over a long horizon. According to Google, the model is built for sustained agentic workflows and handles complex, multi-step problems in software engineering, full-stack web development and scientific research.

Why it matters

The release is aimed at developers who build tools and applications on Google's AI platforms. Google says the model has stronger alignment and reliable follow-through, which it says helps developers build and ship secure products. These are the company's own claims, and the announcement text does not include benchmark figures, test results or comparisons with other models.

Details

Google says developers can access the model by using the API's paid tier, via Google AI Studio and Android Studio. Subscribers to Google AI Ultra can try it in AI Studio and in Google Antigravity, which Google describes as its agent-first development platform. For model details and setup instructions, Google refers developers to the Gemini API Developer Guide. Information about using the model inside Antigravity is in that platform's own blog. Pricing was not announced in the text, and no technical specifications such as context size were given.

Key facts