← The feed Many Minded · The brief
KNOWLEDGE · forward · impact 2/5 · 2026-11-15 · Google DeepMind

Google releases Gemini 4 Hydrogen, built for fast coding and agent work

Google DeepMind's Hydrogen model targets quick coding and agentic tasks at $1.50 per million input tokens. The performance figures are Google's own.

Google DeepMind announced Gemini 4 Hydrogen on November 12, 2026, in a post by Koray Kavukcuoglu, Google's Chief AI Architect. Hydrogen sits below Gemini 4 Carbon and Gemini 4 Argon, the frontier models for deep, long-horizon reasoning. Hydrogen is aimed at fast execution in coding and agentic workflows. It is available in all tiers of Google AI Studio, the Gemini API, Gemini for business, the Gemini app and Google Search. Paid Google AI Pro and Ultra subscribers get it in Google Antigravity.

Pricing is flat across the full 1 million-token context window: $1.50 per million input tokens and $7.50 per million output tokens. The model supports up to 1 million output tokens and has four thinking levels: low, medium, high (the default) and a new xhigh. On the Artificial Analysis Intelligence Index it scores 51 at the default setting, against 50 for Argon, and 54 at xhigh. Against Gemini 3.8 Flash, Google reports tasks finishing 2.5x faster, 57% versus 18.2% on Terminal-Bench 4.1, and more than 5x fewer fabricated claims. Google also launched Orchestrator in Antigravity, which switches between models such as Carbon and Hydrogen within one session. On Terminal-Bench 4.0 it reached near-Carbon success with 24% less quota and 19% faster completion on average.

Why it matters for abundance: knowledge work such as coding, tax and legal research, and data queries is increasingly done by agents. This release puts a model scoring near Argon on one index into consumer, developer and enterprise products at a published, flat rate. A single price across the whole context window makes the cost of long-document work easier to predict. Hydrogen also upgrades Google Search with better intent understanding and more visual answers.

Caveats: every performance and safety figure is Google's own, and many come from internal or partner testing. Some comparisons are to unnamed competing models in its class. Independent evaluation and real-world pricing in practice are what to watch next.

Source: Google DeepMind