Budget for the price after November
Google opened Gemini 4 Carbon to paying developers and enterprises on November 10. For teams evaluating it through the API, the November price is temporary. Google lists introductory charges of $2 per million input tokens and $10 per million output tokens. It says those rates become $4 and $20, respectively, after the introductory period expires on November 30, 2026.
The rates apply to the model Google is making publicly accessible. At the stated base rates, one million input tokens and 100,000 output tokens would cost $3 during the introduction and $6 afterward. That arithmetic excludes any other charges or discounts. A pilot that continues into December should account for the higher rates before measuring its cost per completed task.
Access depends on the product
Google lists the paid Gemini API and Gemini for business as developer and enterprise routes. It separately describes a rollout to Google AI Ultra subscribers in the Gemini app and AI Mode in Search. The announcement does not give a country-by-country launch list for Carbon. CGH did not verify access in any individual account or market.
What the model card establishes
Google’s model card describes text output from text, image, audio or video inputs, with a combined context limit of up to one million tokens. That limit describes how much material the model can handle in one context. The time, token use and reliability of a multi-step job will depend on the workload. CGH reviewed Google’s documents without running API calls or reproducing performance claims.