Google Releases Gemini 4 Hydrogen, Which Outscores Argon on One Benchmark

What happened

Google has released Gemini 4 Hydrogen, a fast general-purpose model for developers and everyday users. The Decoder reports that Hydrogen scores 51 on the Artificial Analysis Intelligence Index at its default high thinking level, versus 50 for the larger Gemini 4 Argon, and reaches 54 at xhigh. Google says Hydrogen is available globally across its platforms; the article also notes weaker results on terminal work than Claude Sonnet 5.5 and GPT-6.1 Sol.

Read next

• More changes in today’s briefingGoogle DeepMind Introduces Gemini 4 Hydrogen for Fast Agentic Workflows

Reported by 11 sources

The story so far

Google launched Gemini 4 Hydrogen across Search, the Gemini app, Google AI Studio, the Gemini API, Gemini for business, and Google Antigravity for AI Pro and Ultra subscribers; Cursor also made it available. Google says Hydrogen completes tasks 2.5 times faster than Gemini 3.8 Flash. It is the default model in Search AI Mode and handles the hardest questions in AI Overviews, where Google says it brings more links, visuals, structured responses, and comparison tables. Access depends on the product: Google says Hydrogen is available across all AI Studio and Gemini API tiers, with usage limits on free tiers, while AI Pro and Ultra subscribers can use it in Antigravity. A report says Gemini app subscribers on Plus, Pro, and Ultra can select Hydrogen, while free users may get it through automatic model selection. Cursor reports a 54.3% CursorBench score at default effort, compared with 39.6% for Gemini 3.8 Flash. Artificial Analysis ranked Hydrogen fourth overall on its AA-Robotics v0.1 benchmark, the highest-ranked Google model.