# Gemini 4 Hydrogen: price, free access and when to use it

> Google released Gemini 4 Hydrogen on 2026-11-12: it is free in Google Search and on the API free tier, and free users of the Gemini app get it through auto model selection, when it fits the prompt; on the paid API, it costs $1.50 per million input tokens and $7.50 per million output tokens, twice the price of Gemini 3.8 Flash.

- Published: 2026-11-13
- Publisher: Rig Labs (https://riglabs.ai/)
- Canonical: https://riglabs.ai/blog/gemini-4-hydrogen/
- Answers: gemini 4 hydrogen
- Agent file (a short card for your agent): https://riglabs.ai/blog/gemini-4-hydrogen/AGENT.md

---

You can now use a Gemini 4 model for free: Google released Gemini 4 Hydrogen on November 12, 2026, in Google Search for everyone, in the Gemini app (free users through auto model selection) and on the Gemini API free tier, per [Google DeepMind's announcement](https://deepmind.google/blog/introducing-gemini-4-hydrogen/). Hydrogen is the fast, lower-cost member of the Gemini 4 family, generally available, and Gemini app and free-tier users pay nothing extra.

On the paid API it costs $1.50 per million input tokens and $7.50 per million output tokens, twice Gemini 3.8 Flash ($0.75/$3.75) and 25% below Gemini 4 Carbon and Argon ($2/$10). That price matters only to people who pay for the API.

## Where can you use Gemini 4 Hydrogen?

Gemini 4 Hydrogen is available in these places, per Google:

- **Google Search**: for all users.
- **Gemini app**: in the model drop-down. Free users get it through auto model selection, when it fits the prompt. Google says Hydrogen follows your tone and rules more closely, and calls the right tools more often across Gmail, Google Drive and connected apps such as Airtable and Dropbox.
- **Google AI Studio and the Gemini API**: all tiers, model name `gemini-4-hydrogen`.
- **Gemini for business.**
- **Google Antigravity**, Google's coding agent app: for paid Google AI Pro and Ultra subscribers.
- **Cursor**: added the same day, [per Cursor](https://x.com/cursor_ai/status/2120944639633211619).

## How much does Gemini 4 Hydrogen cost?

Gemini 4 Hydrogen costs the same across its whole 1 million-token context window. The [Gemini API pricing page](https://ai.google.dev/gemini-api/docs/pricing) lists:

| Model | Input, per 1M tokens | Output, per 1M tokens |
|---|---|---|
| Gemini 4 Hydrogen, free tier | Free | Free |
| Gemini 4 Hydrogen, paid tier | $1.50 | $7.50 |
| Gemini 3.8 Flash, through Dec 31, 2026 | $0.75 | $3.75 |
| Gemini 4 Carbon and Argon, through Nov 30, 2026 | $2.00 | $10.00 |
| Gemini 4 Carbon and Argon, from Dec 1, 2026 | $4.00 | $20.00 |

Simon Willison [calls the Gemini 4 pricing confusing](https://simonwillison.net/2026/Nov/12/llm-gemini/): Carbon and Argon cost the same as each other, and the gap with Hydrogen changes when their price doubles on December 1. In the Gemini app, Hydrogen comes with your plan.

## What is Gemini 4 Hydrogen good at?

Gemini 4 Hydrogen is built for quick, everyday work: coding and debugging, agent tasks, deep research and everyday writing. Google's numbers:

- It finishes tasks 2.5 times faster than Gemini 3.8 Flash when Hydrogen runs at its default thinking level, high.
- It scores 51 on the Artificial Analysis Intelligence Index, against 50 for Gemini 4 Argon. At `xhigh`, the highest of its four thinking levels, Hydrogen scores 54, against 50 for Gemini 4 Argon.
- It has four thinking levels: low, medium, high (the default) and a new `xhigh`. Lower levels answer faster and cost less.

Google still positions Carbon and Argon for the hardest problems. In Antigravity, a new Orchestrator switches between Carbon and Hydrogen in one session; Google reports that the Orchestrator succeeds on Terminal-Bench 4.0 almost as often as Carbon alone, and uses 24% less of a plan's usage quota than running Carbon alone. That matters to Antigravity users on Google AI Pro and Ultra, whose plans have a usage quota.

One public test: the X user @noclipepe [had four models build a Chrome dinosaur game](https://x.com/noclipepe/status/2121219775170761044). Hydrogen finished in 18 minutes 41 seconds, against 45 minutes 30 seconds for Claude Fable 5.2.

## Should you switch to Gemini 4 Hydrogen?

Gemini 4 Hydrogen is the obvious pick if you used Gemini 3.8 Flash and want a better answer, and if the speed is worth twice the paid price to you. If cost matters most, 3.8 Flash stays cheaper until the end of 2026. For the hardest reasoning, Google still points to Carbon and Argon.

## What to do today

If a script or automation of yours calls Gemini 3.8 Flash, run one real job with the model name changed:

```
gemini-3.8-flash  ->  gemini-4-hydrogen
```

Compare the answer, the time and the token count. Keep Hydrogen where it saves you a retry or a review.

Last checked 2026-11-13.

Sources:
- [Google DeepMind: Introducing Gemini 4 Hydrogen, 2026-11-12](https://deepmind.google/blog/introducing-gemini-4-hydrogen/)
- [Gemini API changelog, 2026-11-12](https://ai.google.dev/gemini-api/docs/changelog)
- [Gemini API pricing](https://ai.google.dev/gemini-api/docs/pricing)
- [Simon Willison: llm-gemini 0.37, 2026-11-12](https://simonwillison.net/2026/Nov/12/llm-gemini/)
- [Cursor on X, 2026-11-12](https://x.com/cursor_ai/status/2120944639633211619)
- [@noclipepe on X, 2026-11-13](https://x.com/noclipepe/status/2121219775170761044)
