The short answer On November 10, 2026, Google opened Gemini 4 Carbon to paying developers and enterprises and began rolling it out to Google AI Ultra subscribers. Introductory API prices of $2/$10 per million tokens run through November 30, then rise to $4/$20 on December 1.
Rows of servers in a data center, representing Gemini 4 Carbon's expanded API and enterprise access

What changed

Gemini 4 Carbon first rolled out on October 28 to a limited set of cybersecurity defenders. On November 10, Google expanded access to paying developers and enterprises, positioning Carbon for complex coding, scientific research, cybersecurity, and multistep AI workflows[1][2].

Developers can now call Carbon through the Gemini API in AI Studio and Android Studio, while enterprises access it through Gemini for business. Google AI Ultra subscribers are also getting Carbon access as part of the rollout[2].

API pricing and the December 1 increase

Carbon's introductory API rate is $2 per million input tokens and $10 per million output tokens, including tokens used for reasoning. Those rates run through November 30, then double[2][3].

Token typeThrough Nov 30From Dec 1
Standard input$2$4
Standard output$10$20
Cached input$0.10$0.20

Carbon supports implicit caching but not explicit caching, and separate pricing applies to Batch, Flex, and Priority processing[3]. The rates match Gemini 4 Argon's published pricing, including the upcoming December increase[3].

Enterprise and developer access

Google Cloud's release notes confirm Carbon's availability in Gemini Enterprise's web and mobile apps, Workflow Builder, and AI developer tools. Administrators must enable Carbon for web and mobile users, while organizations using custom authorized-model lists must approve it separately[3].

Separately, GitHub added Gemini 4 Carbon to GitHub Copilot for eligible paid plans, including Business and Enterprise, with administrator controls and separate usage-based billing[5].

What it actually costs to run

Because Carbon can consume additional billable output tokens during extended reasoning, per-token parity with Argon does not guarantee identical per-task cost. An eWeek calculation shows a 10M-input + 1M-output workload costing $30 through November, rising to $60 from December 1[3].

Verdict: The November window is a useful trial period, but treat it as a testing opportunity rather than evidence of long-term cost efficiency. Before routing production workloads to Carbon, measure reasoning-token consumption and completed-task cost at December rates[3].

Frequently asked questions

Is Gemini 4 Carbon available to everyone now?

No — general availability is for paying developers, enterprises, and Google AI Ultra subscribers. Google's specialized "Carbon for cyber" remains restricted through its Fairwind Program[2][3].

How much does Gemini 4 Carbon cost?

$2 per million input tokens and $10 per million output tokens through November 30, rising to $4 and $20 on December 1[3].

Where can I access Gemini 4 Carbon?

Via the Gemini API in AI Studio and Android Studio, Gemini for business, and GitHub Copilot on eligible paid plans[2][5].

Sources

  1. Google Blog — "Gemini 4 Carbon now available" — blog.google
  2. Unite.AI — "Google Expands Gemini 4 Carbon Access Beyond Cybersecurity Defenders" — unite.ai
  3. eWeek — "Google Opens Gemini 4 Carbon to Businesses, but API Prices Double Dec. 1" — eweek.com
  4. Google AI for Developers — Gemini API pricing — ai.google.dev
  5. GitHub Changelog — "Gemini 4 Carbon in GitHub Copilot" — github.blog