Google Releases Gemini 4 Carbon Frontier Model and Announces Pricing
Understand the story
On November 11, 2026, Google DeepMind released and fully launched the frontier model Gemini 4 Carbon. The model is positioned as the strongest model for high-intensity tasks, with use cases including cybersecurity, long-horizon agent workflows, complex multi-step coding, open-ended research, and scientific reasoning. According to The Verge, the model was opened to some cybersecurity defenders two weeks before its public launch; meanwhile, Google also launched the swarm agent feature in the Antigravity preview. Regarding pricing, the API price is $2 per million input tokens and $10 per million output tokens, the same as Gemini 4 Argon. The Verge called this introductory pricing, roughly one-fifth of GPT-6 Astra's and one-quarter of Fable 5.2's, and prices will double after the introductory period ends at the end of November, though still lower than competitors. Previous versions were thought to have no follow-up reports; The Verge's report adds information on the early access for cybersecurity defenders, the Antigravity swarm feature, and the upcoming price increase after introductory pricing, while both reports are consistent on the basic facts of the model's official release and pricing.
AI overview of reporting · Updated 47 min前
Reporting timeline
Explore different aspects of the story through its reporting.
- Google DeepMindSelectedGemini 4 Carbon is now available
Google DeepMind has fully launched Gemini 4 Carbon, aimed at demanding tasks such as cybersecurity, long-horizon agent workflows, complex multi-step coding, and scientific reasoning. API pricing matches Gemini 4 Argon ($2 per million input tokens, $10 per million output tokens).
- The Verge · AISelectedGoogle releases Gemini 4 Carbon with agent ‘swarm’ feature
Google has made Gemini 4 Carbon publicly available after two weeks of limited access for some cybersecurity defenders, positioning it as its strongest model for demanding tasks like complex multi-step coding and open-ended research. Introductory pricing is $2 per million input tokens and $10 per million output tokens, roughly one-fifth the cost of GPT-6 Astra and one-quarter that of Fable 5.2; prices will double after the introductory period ends in late November, still below competitors.
Story heat trend
Not enough continuous observations to show a trend yet.