Gemini 4 Carbon: long jobs, small windows

I don’t want a long-running model to turn a small desk screen into a scrolling transcript. Carbon makes me think about a different interface: a compact view of the job, its last verified checkpoint, and the next decision that needs me.

What Google announced

Google made Gemini 4 Carbon available on 10 November, describing it as a model for long-horizon work, including complex coding and research. The announcement says it uses more tokens than Argon for deeper exploration and verification. Those are the provider’s descriptions, not results from my own testing.

The announced introductory API price matches Argon: $2 per million input tokens and $10 per million output tokens. The introductory period expires on 30 November 2026; the same post gives the subsequent public-access Carbon price as $4 input and $20 output per million tokens.

What belongs on the desk

My proposed display would show a task name, elapsed time, the last confirmed action, and whether input is required. A short voice session could ask for a checkpoint without pretending the underlying job is finished. This is an interface direction, not a working Mosaico integration.

A bounded experiment

I’d give it one synthetic debugging task with a fixed budget and known checks. Then I’d interrupt and resume the job, recording what survives and whether the status display remains accurate. I would compare the whole task cost at both introductory and later prices. I have not run that experiment.

Official source, published 10 November 2026: Gemini 4 Carbon — read the original announcement/research ↗. Source checked 12 November 2026.