- Google announces “Gemini 4 Hydrogen,” a new highly efficient and fast AI model from the next-generation Google AI model family, “Gemini 4.”
- A major upgraded AI model succeeding “Gemini 3.8 Flash.”
- Rollout has begun globally and sequentially through the “Gemini” app’s model selection dropdown and “Google Search.”
On Thursday, November 12, 2026 (local time), Google announced “Gemini 4 Hydrogen,” a new highly efficient and fast AI model from the next-generation Google AI model family “Gemini 4,” and made it available for developers, enterprises, and general users.
“Gemini 4 Hydrogen” is a major upgraded AI model succeeding “Gemini 3.8 Flash.” Unlike the frontier models in the same “Gemini 4” series, such as “Gemini 4 Carbon” and “Gemini 4 Argon,” which specialize in deep reasoning and complex problem-solving, “Gemini 4 Hydrogen” is designed to execute everyday coding and diverse agentic workflows quickly and efficiently.
At the default “thinking level” option of “high,” it completes tasks 2.5x faster than “Gemini 3.8 Flash” while scoring more than three times higher on the “Terminal-Bench 4.1” coding benchmark (57% vs. 18.2%). Furthermore, on the “Artificial Analysis Intelligence Index,” it achieves a higher score than “Gemini 4 Argon” under default settings (51 vs. 50, and 54 at the maximum “xhigh” setting).
In addition, a new “xhigh” thinking level option has been added alongside “low,” “medium,” and “high” (default). With a total of four levels available, users can flexibly adjust processing speed and cost according to the depth of reasoning required for the task.
The development environment “Google Antigravity” has also introduced a new feature called “Orchestrator,” which automatically and dynamically switches between models such as “Gemini 4 Carbon” and “Gemini 4 Hydrogen” within a single session. On “Terminal-Bench 4.0,” this reduces quota consumption by an average of 24% and increases completion speed by 19% while maintaining task success rates close to those of using “Gemini 4 Carbon” alone.
“Gemini 4 Hydrogen” is now available for the Gemini API via Google AI Studio, Android Studio (paid plans), and Gemini for Business, with flat pricing across the entire 1 million token context window ($1.50 per million input tokens / $7.50 per million output tokens). It is also available for Google AI Pro and Google AI Ultra members in Google Antigravity and Gemini features within Google Sheets.
For free general users, the global rollout has begun sequentially through the “Gemini” app’s model selection dropdown and “Google Search.”
Source: Google (https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-hydrogen/)