GooglemodelGemini 4 Carbon (Generally Available)
Google's strongest reasoning model is now generally available on the paid Gemini API tier, and rolling out in the Gemini app to Google AI Ultra subscribers.
Details
- Model id is gemini-4-carbon. It spends more thinking tokens per turn than Gemini 4 Argon to test hypotheses and check its own work, so it is slower and uses more tokens. Argon stays the faster pick for everyday coding.
- One 1,048,576-token window is shared by input, thinking, and output. Thinking levels are low, medium (the default), and high. Answers longer than 65,536 tokens need background execution.
- Introductory API pricing is $2.00 per 1M input tokens and $10.00 per 1M output tokens through November 30, 2026, then $4.00 and $20.00 from December 1.
- API access is on the paid tier of the Gemini API and Google AI Studio, not the Free tier. Explicit context caching is not supported.
- In the Gemini app it is rolling out to Google AI Ultra subscribers on desktop, mobile, and web.
Best for
Long, hard agentic work: complex coding, dense PDFs and spreadsheets, science, and cybersecurity.
When to use it
The faster Gemini 4 Argon is not getting the job done and you can wait minutes for a deeper answer.
Who benefits
Developers building agents, and Ultra subscribers using Gemini on hard projects.
Join the newsletter to get ahead with AI.
I share strategies, new modules, free guides, and tools to get you better at using AI in your work.
Leading a team? See team programs →