Google released Gemini 4 Hydrogen on November 12, 2026, and put it straight into Google Search for all users. Google calls it the fast, lower cost model in the Gemini 4 family, built for AI agents that run many quick steps.
For developers, the headline is price. Hydrogen costs $1.50 per million input tokens and $7.50 per million output tokens, flat across its full 1 million token context window.
Key facts
- API price: $1.50 input and $7.50 output per million tokens.
- 1 million token input, up to 1 million output tokens.
- Live in AI Mode in Google Search for all users, in every supported AI Mode language.
What Google says Hydrogen does
Gemini 4 Carbon and Gemini 4 Argon remain Google's frontier models for long, deep reasoning. Hydrogen sits below them and aims at fast execution. It has four thinking levels: low, medium, high (the default) and a new xhigh.
Google's own benchmark figures include:
- 51 on the Artificial Analysis Intelligence Index at high thinking, one point above Argon's 50, and 54 at xhigh.
- 57% on Terminal-Bench 4.1, against 18.2% for Gemini 3.8 Flash.
- 52.3% on the Vals Tax Agent Bench, which Google lists as first place.
Google also says Hydrogen finishes tasks 2.5 times faster than Gemini 3.8 Flash at high thinking. These are Google's numbers, not independent tests.
Who can use it, and what it costs
- Google Search: AI Mode uses it for all users, globally.
- Gemini app: in the model menu. Free users get it when automatic model selection picks it.
- Google Sheets: Gemini in Sheets gets it first for Google AI Pro and Ultra subscribers, plus eligible business plans.
- Developers: Google AI Studio on all tiers, Android Studio and the Gemini API.
A worked example with Google's rates. Say your website chatbot sends 10 million input tokens and gets 2 million output tokens back in a month:
| Model | Input | Output | Total |
|---|---|---|---|
| Gemini 4 Hydrogen | $15 | $15 | $30 |
| Gemini 4 Carbon (from Dec 1, 2026) | $40 | $40 | $80 |
The Carbon row uses the US$4 and US$20 rates Google lists after the Carbon launch price ends on November 30. See our Gemini 4 Carbon story.
What it means for a small business
Your customers now see Hydrogen's answers in AI Mode when they search for what you sell. That makes clear, current facts on your own site matter more: opening hours, prices, service areas. If a developer built you a chatbot or an automation on Gemini, ask whether Hydrogen can handle the job at a lower bill than a frontier model.
Our take
Based on Google's announcement, not our own testing: Hydrogen is the Gemini 4 model most owners will meet, because it runs in Search and in the free app. For API work, the flat price across the whole context window makes costs easier to forecast. Run your own test on your real tasks before you switch, since the benchmark claims come from Google.
Frequently asked questions
Is Gemini 4 Hydrogen free?
Free Gemini app users get it when automatic model selection chooses it, and AI Mode in Search uses it for all users. API use costs $1.50 per million input tokens and $7.50 per million output tokens.
How big is the Gemini 4 Hydrogen context window?
Google lists 1 million tokens of input and up to 1 million output tokens.
Is Hydrogen better than Gemini 4 Carbon?
Google positions Carbon and Argon as its frontier models for deep reasoning, and Hydrogen as the faster, cheaper model for agent work. Google reports Hydrogen scoring one point above Argon on one index.



