Google Launches Gemini 4 Hydrogen in Search, the Gemini App and API
Gemini 4 Hydrogen is Google's fast Gemini 4 model, now in Search, the Gemini app and the API. What changes, the price, the four thinking levels.
Google launched Gemini 4 Hydrogen on 12 November, the fast, everyday model of its Gemini 4 family. It's rolling out globally to everyone: in Google Search for all users, in the Gemini app's model menu, in the Gemini API and Google AI Studio, and for paid Google AI Pro and Ultra subscribers in Google Antigravity, its coding tool.
Google positions Hydrogen next to its two frontier models, Gemini 4 Carbon and Gemini 4 Argon, which handle long, deep reasoning. Hydrogen is built for quick execution: writing code, running tests, research tasks and everyday questions.
What changes in Google Search
Google calls Hydrogen the biggest model upgrade to Search yet. In Search it brings:
- Better understanding of what you mean: questions with several parts, typos or very specific requests are handled in your natural flow of thought.
- More visual answers: more images, structured responses, comparison tables and custom tools.
- A wider range of links: Search runs more simultaneous queries across the subtopics of your question, so it surfaces sites you might not have found before.
- Better answers in every language supported by AI Mode.
What changes in the Gemini app
In the Gemini app, Google says Hydrogen follows your instructions more closely across longer conversations:
- drafts, study guides and rewrites stick to your tone, format and rules, so they're usable on the first try;
- it breaks a goal into steps and pulls context from your uploads, Gmail, Google Drive and connected apps such as Airtable and Dropbox;
- it calls the right tools more often and builds working prototypes, like interactive pages, in fewer prompts.
Gemini in Google Sheets also gets Hydrogen, starting with Google AI Pro and Ultra subscribers.
For developers: price, speed and four thinking levels
- Price: $1.50 per million input tokens and $7.50 per million output tokens, flat across the full 1 million-token context window. It can also output up to 1 million tokens.
- Thinking levels: low, medium, high (the default) and a new xhigh, so you only pay for the reasoning depth a task needs.
- Against Gemini 3.8 Flash: Google says Hydrogen finishes tasks 2.5 times faster at its default setting, scores 57% on Terminal-Bench 4.1 against 18.2%, and makes more than five times fewer fabricated claims.
- Against Argon: on the Artificial Analysis Intelligence Index, Google reports 51 for Hydrogen at its default level against 50 for Argon, and 54 at xhigh.
Google Antigravity also gets a new Orchestrator that switches between Carbon and Hydrogen within the same session. On Terminal-Bench 4.0, Google says it came close to Carbon's success rate while using 24% less quota and finishing 19% faster on average.
Safety
Google says Hydrogen reduces overreach in agentic tasks by up to 81%, cuts misaligned actions and task shortcuts by 12 times compared with 3.8 Flash, and is more than 2.3 times more resistant to indirect prompt injection on the Gray Swan benchmark. All benchmark figures are Google's own.
Gemini is also coming to your face: Samsung's Gemini glasses launch on 10 December from $395.
