Google ships Gemini 4 Hydrogen to everyone, free tier included, and its fast model edges past last month's flagship

Google on Nov. 12 released Gemini 4 Hydrogen, a fast, cheap everyday model, across every tier of the Gemini app, Google Search and the API.

Google says Hydrogen scores 51 on the Artificial Analysis Intelligence Index at its default high setting, ahead of last month's flagship Gemini 4 Argon (50), and 54 at xhigh. It finishes tasks 2.5 times faster than Gemini 3.8 Flash and cuts fabricated claims by more than five times, the company said. Pricing is flat across the full 1-million-token context: $1.50 per million input tokens and $7.50 per million output.

It is live now in the Gemini app including the free tier, Google AI Studio, the Gemini API and Search, and for paid AI Pro and Ultra users in Antigravity. Cursor added it the same day and said it scores 54.3% on CursorBench at default effort, up from 39.6% for 3.8 Flash. On Arena's Agent Arena it debuted at No. 5, with a median of 5.3 minutes per task, three times faster than No. 1 Claude Fable 5.2. Google also added an Orchestrator to Antigravity that hands deep design work to Carbon or Argon and code-and-test work to Hydrogen.

Box CEO Aaron Levie said Hydrogen was 5 points more accurate than 3.8 Flash in early tests, calling it "a serious jump for fast models." Under YouTuber Paul J Lipsky's test video (42,000 views), a top comment read: "Finally, an AI model we can actually use today rather than just reading hype tweets about it." Carbon, announced Oct. 28, took two weeks to reach general availability; Hydrogen shipped on launch day. Some pushed back on Google charting a Flash-class model against frontier models: "look at what they are comparing it to."