Gemini 4 Hydrogen is a Google model built to carry out agent tasks quickly, and the model that brought Gemini 4 to all Google AI Studio tiers, the Gemini app and Google Search.
In plain words
Google's larger Gemini 4 models, Carbon and Argon, handle deep reasoning on the hardest problems. Hydrogen is the quick one: it can work alone on everyday jobs or serve as the worker while Carbon coordinates. It reads up to 1 million tokens and can write as many. Four thinking levels, from low up to a new xhigh, trade quality against speed and cost. Developers pay $1.50 for each million tokens sent in and $7.50 for each million it writes, whatever the context length.
A quick example
Imagine a team splitting a large code change. Gemini 4 Carbon acts as the coordinator and divides the job, while several Hydrogen agents build and test the parts side by side, each in its own workspace.
Where you'll meet it
- Everyday coding and agent work through the Gemini API.
- The Gemini app and Google Search.
- Worker agents paired with Gemini 4 Carbon in Google Antigravity.
Related
Sources
- DocsIntroducing Gemini 4 Hydrogen, Google · read 12 Nov 2026
- DocsGemini 4 Hydrogen, Gemini API docs · read 12 Nov 2026
- DocsGemini 4 Hydrogen and Orchestrator, Google Antigravity · read 12 Nov 2026
Checked by WhatIsAI against the sources above · 12 Nov 2026. Spotted an error? Report it.