19 November 2026 · Bucharest · Europe/Bucharest

THE WORLD, in brief

BREAKING Robert Pattinson is officially the next James Bond, Amazon MGM announces

Share this article

Share on Facebook (new tab)

Instagram has no direct website sharing link. This opens your device share menu where supported; Instagram may not appear. Otherwise copy the link. Nothing is posted automatically.

A.I. News · 08:30 · 2026-11-13

Google launches Gemini 4 Hydrogen for everyday and agentic work

Google introduced Gemini 4 Hydrogen, a faster model for everyday coding and agentic workflows, and said it is rolling out across the Gemini app, Search, the Gemini API and business tools.

Google launches Gemini 4 Hydrogen for everyday and agentic work
A Google sign on Charleston Road in Mountain View, California (file image). Photo: Dietmar Rabich / Wikimedia Commons, CC BY-SA 4.0, cropped. Source: https://commons.wikimedia.org/wiki/File:Mountain_View_(CA,_USA),_Charleston_Road,_Google-Schild_--_2022_--_2896.jpg / Licence: https://creativecommons.org/licenses/by-sa/4.0/

Google on 12 November introduced Gemini 4 Hydrogen, a new model it said was built for speed and everyday coding and agentic workflows, and is being made available more broadly than its frontier Gemini 4 models.

In a post on the Google blog, Koray Kavukcuoglu, senior vice-president at Google DeepMind and the company's chief AI architect, wrote that while Gemini 4 Carbon and Gemini 4 Argon were designed for deep, long-horizon reasoning, Hydrogen was built for rapid execution. He said Hydrogen was coming to all tiers of Google AI Studio, the Gemini API, Gemini for business, the Gemini app, Google Search, and paid Google AI Pro and Ultra subscribers in Google Antigravity.

Google said Hydrogen surpassed Argon on the Artificial Analysis Intelligence Index at its default "high" thinking level, with a score of 51 against 50, and reached 54 at its highest "xhigh" setting, while often finishing tasks faster than models in its class. The company priced the model at $1.50 per million input tokens and $7.50 per million output tokens across a 1-million-token context window.

Compared with Gemini 3.8 Flash, Google said Hydrogen finished tasks 2.5 times faster at its high default setting, more than tripled its Terminal-Bench 4.1 score, from 18.2% to 57%, and cut fabricated claims by more than five times. The company also introduced Orchestrator in Google Antigravity, a feature that switches between models such as Carbon and Hydrogen in the same session. On Terminal-Bench 4.0, Google said Orchestrator reached near-Carbon success rates while using 24% less quota and finishing 19% faster on average.

For consumers, Google said Hydrogen would improve everyday help in the Gemini app and in Search, including better instruction-following and more visual responses. In the Gemini app, free users would get access through automatic model selection when the system judged Hydrogen to be the right model for a prompt, the post said.

Google said Hydrogen included the latest Gemini 4 safety and security stack and reported gains against indirect prompt-injection attacks and in how the model handles blocked or impossible tasks. Further technical detail is in the Gemini 4 Hydrogen model card published by Google DeepMind.


Source: Google

← Back to A.I. News · Home