Google introduced Gemini 4 Hydrogen on November 12, 2026, as a model for everyday coding and agentic workflows. The release gives teams a Gemini 4 option for routine implementation and testing, including when they want to reserve a larger model for hard design decisions.
How Hydrogen fits into Gemini 4
In its launch announcement, Google presents Hydrogen as an execution model that can work on its own or alongside other Gemini models. Google describes Gemini 4 Carbon as an option for architectural decisions or difficult bugs. It positions Gemini 4 Argon for deeper reasoning as well, while Hydrogen is intended for the repeated work of reading files, making edits and checking results.
Thinking levels and API cost
Hydrogen has four thinking levels: low, medium, high and xhigh. High is the default. Google reports an Artificial Analysis Intelligence Index score of 51 at that setting, compared with 50 for Argon, and 54 for Hydrogen on xhigh. Those are aggregate scores. They do not establish which model will do better on any specific repository or coding task.
The company lists a one million token context window with a flat price of $1.50 per million input tokens and $7.50 per million output tokens. That is token pricing; it does not tell teams how many turns or tool calls a repository task will take.
Where developers can use it
Google says it is rolling out Hydrogen across Google AI Studio and the Gemini API, as well as to its business and consumer products. Paid Google AI Pro and Ultra subscribers get access in Google Antigravity. Google also introduced an Orchestrator for Antigravity that can switch models within one session. For teams considering that setup, a useful first check is whether it gets a representative change through their existing test suite and review process in fewer steps.
