Google releases Gemini 4 Hydrogen — fast mid-tier model nearly...

Google releases Gemini 4 Hydrogen — fast mid-tier model nearly matching Gemini 4 Argon in intelligence

On November 12, 2026, Google released Gemini 4 Hydrogen — the mid-tier model in the Gemini 4 lineup, adapted from Gemini 4 Carbon for fast agentic tasks and everyday coding. It scores 51 on the Artificial Analysis Intelligence Index (54 on xhigh) compared to 50 for Gemini 4 Argon, solves 57% of Terminal-Bench 4.1 tasks, and operates 2.5 times faster than Gemini 3.8 Flash at a price of $1.50 and $7.50 per million input and output tokens.

Compiled by Sergey KostenchukPublished 2026-11-12Updated 2026-11-12

2026-11-12 Research Google

🤖 Google releases Gemini 4 Hydrogen

The model is adapted from the flagship Gemini 4 Carbon for fast agentic tasks and coding. On the Artificial Analysis Intelligence Index, it scores 51 on high compared to 50 for Gemini 4 Argon, and on Terminal-Bench 4.1, it achieves 57% compared to 18.2% for Gemini 3.8 Flash, while being 2.5 times faster.

🌍 In the Gemini 4 lineup, Carbon and Argon handle deep reasoning, while Hydrogen handles fast execution. The Orchestrator in Antigravity automatically switches models: quota consumption dropped by 24%, and speed increased by 19% without a loss in quality.

👤 In the Gemini app and Google Search, Hydrogen is substituted for free, making responses faster and more accurate. Developers can access gemini-4-hydrogen in AI Studio, the Gemini API, and Antigravity for $1.50/$7.50 per million tokens.

Source 1: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-hydrogen/ Source 2: https://deepmind.google/models/model-cards/gemini-4-hydrogen/

Sources