AI tools · November 13, 2026

Google's Gemini 4 Hydrogen is the fast new brain behind Search — and it's free to try

Google has a new fast AI model, and it is already answering your searches. Gemini 4 Hydrogen, launched November 12, is the execution tier of the Gemini 4 family: built for speed across everyday coding and multi-step agent work, while the frontier models Carbon and Argon handle the hardest reasoning. Google calls it the biggest model upgrade to Search yet, and it is now the default model in AI Mode and the one handling the toughest questions in AI Overviews.

The numbers are Google's own, but they are striking. The company says Hydrogen finishes tasks 2.5 times faster than Gemini 3.8 Flash, the model it succeeds, and more than triples its Terminal-Bench 4.1 score, 57% against 18.2%. On the Artificial Analysis Intelligence Index it edges past the larger Argon model at its default setting, 51 to 50, and reaches 54 at a new xhigh thinking level. It costs $1.50 per million input tokens and $7.50 per million output tokens, flat across its full one-million-token window.

For regular users the change is free and automatic: Hydrogen is rolling out to all Search users globally, with better handling of multi-part questions, typos, and specific requests, plus more visual answers and a wider range of links. Developers can try it in Google AI Studio and the Gemini API, and it is in the Gemini app, Gemini for business, and Google Antigravity for Pro and Ultra subscribers. Independent testing of Google's speed and benchmark claims has not been published yet, so treat the headline figures as the vendor's scorecard until third parties verify them.

Why it matters

Search is where most people meet AI, and Google just swapped the engine underneath it for everyone at once. If your complicated questions suddenly get better answers, this is why — and developers get a faster, cheaper model for agent work in the same release.

Key facts

Launched November 12, 2026, as the fast execution tier of the Gemini 4 family, alongside frontier models Carbon and Argon
$1.50 per million input tokens and $7.50 per million output tokens, flat across a one-million-token context window
Google says it finishes tasks 2.5 times faster than Gemini 3.8 Flash and scores 57% on Terminal-Bench 4.1, versus 18.2% for Flash (Google's own figures)
Beats the larger Gemini 4 Argon on the Artificial Analysis Intelligence Index at its default setting, 51 vs. 50, and reaches 54 at its highest thinking level
Four thinking levels: low, medium, high (default), and a new xhigh tier
Now the default model in Google Search's AI Mode and handles the hardest questions in AI Overviews; also in the Gemini app, AI Studio, the Gemini API, and Antigravity for Pro and Ultra subscribers

Sources

Spot an error? We correct quickly and note it. Contact us via the contact page.

← More AI news