Source : The Verge
MOUNTAIN VIEW, California, November 10, 2026 – Google has begun rolling out its most capable artificial intelligence model, Gemini 4 Carbon, to developers, businesses and some consumers, two weeks after first giving it to a small group of cybersecurity defenders. The release was reported by The Verge and set out in a blog post by Koray Kavukcuoglu, Google’s chief AI architect.
Google describes Carbon as its “strongest model”, with frontier performance in cybersecurity and in long-horizon agentic work, meaning tasks in which an AI system works through many steps on its own. Carbon builds on Gemini 4 Argon, which Google launched about two weeks before Carbon was first announced on October 28, 2026. Google says Carbon “pushes the frontier even further” than Argon, but that Argon “remains the ideal choice for the majority of software engineering and knowledge work”. Carbon, it says, is built for tasks that test the limits of a model’s intelligence, such as complex multi-step coding and debugging and open-ended scientific research.
Alongside the model, Google is previewing a feature called “/swarm” in Google Antigravity, its agent-first development platform. Google says the feature turns a single prompt into a coordinated team of agents that users can track in real time. Developers can assign Carbon to roles that demand the deepest reasoning, such as architectural design, coordination and auditing, while pairing it with faster, lighter models working in parallel, isolated workspaces.
On availability, Carbon is offered through the Gemini API in Google AI Studio and in Gemini for business to all paying developers and enterprises. Google AI Ultra subscribers are receiving it in Google Antigravity, the Gemini app and AI Mode in Search. The Verge noted that it is not clear when, or whether, Carbon will reach lower-tier subscribers or free users.
Pricing is a central part of Google’s pitch. Carbon launches at an introductory price of $2 per million input tokens and $10 per million output tokens, the same as Argon. Google says that period ends on November 30, 2026, after which the price rises to $4 per million input tokens and $20 per million output tokens. The Verge reported that even at the higher rate Carbon would remain considerably cheaper per token than rival frontier models, though it cautioned that this does not necessarily mean lower costs, because different models can use different numbers of tokens for the same task.
Google also cited independent benchmarks. It says Carbon ranks first on CursorBench 4.0, Perplexity’s WANDR and Zapier’s AutomationBench, at more than twice as low a cost per task as the next best external model. When it announced Carbon on October 28, Google said the model led the Artificial Analysis Intelligence Index, version 5.0, with a score of 59. The Verge cautioned that companies can and do choose benchmark results that show their tools in the best light, so the figures are Google’s own claims.
Google has also opened applications for scientists and researchers, who can seek free access to both Argon and Carbon through Antigravity’s highest usage tier for six months, subject to eligibility, capacity and selection. Cybersecurity defenders can apply to the Fairwind Program for access to Carbon for cyber work on a rolling basis.
The release marks a return to the front of the field for Google. For months, frontier AI development had looked like a race between Anthropic and OpenAI, The Verge said, and Google’s two frontier launches in quick succession are aimed squarely at changing that picture.
About The Author
