New to this? Read it in simple words
- Google’s strongest model, Gemini 4 Carbon, is now open to paying users.
- It costs $2 per million input tokens until November 30, then $4.
- A new /swarm preview lets one prompt start a team of AI agents.
- An outside test group scores it 60, behind Claude Fable 5.2 at 65.
- Tokens
- Small pieces of text that AI models read and write, and that services charge for.
- API
- A way for other programs to use a service, such as an AI model, directly.
- Agent
- An AI system that takes steps on its own to finish a task.
Google announced Gemini 4 Carbon on October 28, and gave it first to cyber defenders in its Fairwind Program. Google says Gemini 4 Argon, released last month, remains the better choice for most coding and office work.
Who can use it
Carbon is in the Gemini API, through Google AI Studio, for all paying developers. Businesses get it through Gemini for business. It is not on the API’s free tier.
Google AI Ultra subscribers can use it in the Gemini app, in AI Mode in Search and in Antigravity. It is not clear when, or if, cheaper plans or free users will get it, The Verge reports.
Index scores from Artificial Analysis. Prices from Google.
The price
Until November 30, Carbon costs $2 per million input tokens and $10 per million output tokens, the same as Argon. From December 1, the price doubles to $4 and $20.
Even the lower price is about a quarter of Claude Fable 5.2’s, The Verge notes. But Google says Carbon uses more tokens than Argon, because it explores and checks its work for longer.
A team of agents
/swarm is in preview in the Antigravity app and its command-line tool. One prompt starts a coordinator that hands work to several agents running side by side, each in its own copy of the code.
Other agents act as checkers that test every change, Google says. Developers can give Carbon the hardest roles, such as design and review, and use faster models for the rest.
The tests
Google says Carbon ranks first on three outside tests of real work, including CursorBench 4.0, at more than half the cost per task of the next model. These are Google’s own summaries.
Artificial Analysis, an outside test group, now scores Carbon (high) at 60 on its Intelligence Index, fourth of 74 models. Claude Fable 5.2 (max) leads with 65.
At its first release, Carbon scored 59 and led the index. The test group has updated its index since then, so the two scores are not exactly comparable.
Sources
Every fact in this story comes from the sources below. Open them to check our work.
- 1
- 2
- 3
- 4Research · November 10, 2026Google Expands Gemini 4 Carbon Access to Power Users Worldwide Android Headlines
- 5Research · November 10, 2026Google Gemini 4 Carbon is now publicly available — and it undercuts Claude Fable 5.2 and Opus 5.5 on API price VentureBeat
- 6Research · November 10, 2026Gemini 4 Carbon: Intelligence, Performance & Price Analysis Artificial Analysis
The index score comes from Artificial Analysis, an outside test group. The rankings on other tests are Google’s own.
Google opens Gemini 4 Carbon to paying users. Silicon AI News, November 10, 2026. https://siliconainews.com/stories/google-gemini-4-carbon-wider-release
Google opened its strongest model to paying users at about a quarter of a rival’s price, though it lost its lead on an outside test. How we rate the mood
The mood shows how the news frames the event for the tech and AI world: as an opportunity (6 and above) or as a loss (below 5). It is not our verdict on who is right.