# Google's Gemini 4 Hydrogen is its new everyday model: $1.50/$7.50 per million tokens, 1M context, now powering Search

> Google launched Gemini 4 Hydrogen on November 12 as the fast "daily driver" of the Gemini 4 family. It is rolling out in Search, the Gemini app, Sheets and the API. Google's own benchmarks put it level with Argon on one intelligence index, at a lower list price than Argon or Carbon. Here is what it costs, where it runs and which claims are only Google's.

Published Nov 13, 2026 by TechMeasured Staff · AI · Canonical: https://techmeasured.com/ai/google-gemini-4-hydrogen-1-50-7-50-1m-context-xhigh-thinking-search-default-what-changes

Google released **Gemini 4 Hydrogen** on November 12. It is the third model in the Gemini 4 family, after Argon in October and [Carbon on October 28](/ai/gemini-4-carbon-cyber-defenders-first). Google calls Hydrogen its "daily driver" for agentic work. It's built for speed and everyday coding, and leaves the longest, hardest reasoning jobs to Carbon and Argon. It went out to developers, Gemini app users and Google Search on the same day.

Why it matters: this is the model most people will actually meet. Google's Search team calls the change "the biggest model upgrade to Search yet." For developers, Hydrogen is the clear upgrade path from Gemini 3.8 Flash. By Google's own numbers it scores at Argon's level on one benchmark, at a lower price than either bigger model.

## What it costs

Google's developer post doesn't list prices. Data Studios and Chrome Geek both report the same figures from Google's launch and pricing pages:

| Gemini API (per 1M tokens) | Gemini 4 Hydrogen |
|---|---|
| Input | $1.50 |
| Output | $7.50 |
| Cached input (implicit caching, prompts ≥4,096 tokens) | $0.075 |
| Context window | 1,048,576 tokens (shared by input, thinking and output) |
| Default output cap | 65,536 tokens per call |
| Model ID | `gemini-4-hydrogen` |

Both outlets say the price is flat across the whole context window, with no higher tier for long prompts. Data Studios adds that explicit context caching isn't supported, and it lists no batch price.

Two comparisons matter for budgets. First, Data Studios reports that **Gemini 3.8 Flash** is on an introductory $0.75/$3.75 rate through December 31. It then rises to $1.50/$7.50 on January 1, the same as Hydrogen. Teams on Flash should test Hydrogen before that price change. Second, as we reported, Carbon's introductory rate is $2/$10 through November 30 and rises to $4/$20 on December 1.

## Where it runs

- **Gemini API and Google AI Studio:** available now, including the free tier with rate limits.
- **Android Studio:** available (Chrome Geek and Data Studios say the paid tier).
- **Gemini for business:** through the API or the model drop-down.
- **Google Antigravity:** for paid Google AI Pro and Ultra subscribers. Its new "Orchestrator" mode pairs Hydrogen with Carbon.
- **Gemini app:** rolling out. Paid users can pick it from the model menu. Free users get it when the app's automatic model selection decides it fits the prompt, per Digital Trends.
- **Google Sheets:** rolling out to AI Pro and Ultra first, plus eligible business plans.
- **Google Search:** rolling out globally to all users, in all languages AI Mode supports, according to Google's Search post. 9to5Google reports, via Chrome Geek, that Hydrogen is now the default model for AI Mode.

## What Google claims, in numbers

All of these figures are Google's own. None had been independently reproduced at publication.

| Claim | Figure |
|---|---|
| Artificial Analysis Intelligence Index | 51 at default "high" thinking, 54 at "xhigh" (Argon: 50) |
| Terminal-Bench 4.1 | 57% (Gemini 3.8 Flash: 18.2%) |
| HLE-Diamond | 58.8% |
| Agents' Last Exam v1.1 | 71.6% |
| Vals Tax Agent Bench / Legal Research Bench | 52.3% (first) / 55.8% (tied first) |
| Speed vs Gemini 3.8 Flash | 2.5x faster task completion at default settings |
| Fabricated claims vs 3.8 Flash | more than 5x fewer |

Digital Trends, summarising Google's charts, says Hydrogen beats OpenAI's GPT-6.1 Terra and Anthropic's Claude Sonnet 5.5 on knowledge benchmarks and leads on image and video tasks. It trails them slightly on science and math, computer use and Terminal-Bench 4.1. Google also says Hydrogen cuts "overreach" in agentic tasks by up to 81%. It says Hydrogen is up to 2x less likely to work around a permission boundary when a task is blocked.

Hydrogen has four thinking levels: low, medium, high (the default) and a new **xhigh**. Chrome Geek notes that Carbon and Argon default to medium and top out at high. So Hydrogen at xhigh can think longer per request than the bigger models, and it bills those extra output tokens at its own lower rate.

## What to do

- **On Gemini 3.8 Flash in production:** run your evals on `gemini-4-hydrogen` now. The per-token price will match Flash in January, and Google claims large speed and accuracy gains. Treat the 2.5x speed figure as a claim until your own latency numbers back it up.
- **Watch the thinking level.** "high" is the default. If cost matters more than depth, set "low" or "medium" explicitly. Only use "xhigh" where you've measured that it helps.
- **Publishers and SEO teams:** Google says Hydrogen in Search fans out into more sub-queries and shows a wider range of links. Expect AI Mode traffic patterns to shift over the coming weeks as the rollout reaches more users.
- **Not confirmed yet:** Data Studios notes Google's posts don't mention Vertex AI or the Gemini CLI. Check Google Cloud's model list before planning around it there.


## Sources

1. [Google — Gemini 4 Hydrogen: everyday coding and agentic workflows for developers](https://blog.google/innovation-and-ai/technology/developers-tools/build-with-gemini-4-hydrogen/)
2. [Google — Google Search gets a major upgrade with Gemini 4 Hydrogen](https://blog.google/products-and-platforms/products/search/gemini-4-hydrogen-search-upgrade/)
3. [Digital Trends — Google's next-gen Gemini 4 Hydrogen AI model is here](https://digitaltrends.com/cool-tech/googles-next-gen-gemini-4-hydrogen-ai-model-is-here-touting-faster-intelligence-for-all)
4. [Data Studios — Gemini 4 Hydrogen: $1.50/$7.50 API pricing, 1M context, Gemini app and Search rollout](https://www.datastudios.org/post/google-gemini-4-hydrogen-launch-api-pricing-gemini-app-search)
5. [Chrome Geek — Gemini 4 Hydrogen Is Out: Who Gets It and What It Costs](https://www.chromegeek.com/gemini-4-hydrogen/)
