Google launched Gemini 4 Hydrogen on November 12, a fast, cheaper Gemini 4 model built for everyday coding and agentic workflows — and put it straight into Google Search for every user worldwide. Why it matters: Hydrogen is priced at $1.50 per million input tokens and $7.50 per million output tokens, and Google says it beats its own flagship Argon on a broad intelligence index at its default setting, which makes it the model most people will actually end up using.
November 12, 2026
What happened
Google introduced Gemini 4 Hydrogen as the speed-and-efficiency member of the Gemini 4 family. Where Gemini 4 Carbon and Gemini 4 Argon are pitched at deep, long-horizon reasoning on the hardest problems, Hydrogen is built for rapid execution across many kinds of agentic work. It replaces the old “Flash” branding, according to 9to5Google, and it is rolling out globally from today across all tiers of Google AI Studio, the Gemini API, Gemini for business, the Gemini app, Google Search, and Google Antigravity for paid Google AI Pro and Ultra subscribers. Free users reach it in the Gemini app through automatic model selection, Google says.
The numbers Google is claiming
Hydrogen offers a native 1 million-token input window and up to 1 million output tokens, priced flat across the whole context with no long-context surcharge. It has four thinking levels — low, medium, high (the default) and a new extra-high “xhigh” tier. On the Artificial Analysis Intelligence Index, Google says Hydrogen scores 51 at its high default, one point above Argon’s 50, and 54 at xhigh. Against Gemini 3.8 Flash, its predecessor, Google claims tasks finish 2.5 times faster at the high setting, Terminal-Bench 4.1 scores more than triple (57% versus 18.2%), and fabricated claims drop by more than five times. Those are Google’s reported figures, not independent results. Google also says Hydrogen ranks fifth on Agent Arena for multi-step work and leads its class on HLE-Diamond at 58.8%.
Orchestrator: Carbon thinks, Hydrogen runs
Alongside the model, Google introduced Orchestrator in Antigravity, which switches between models inside a single session instead of using one model start to finish — Carbon can coordinate while Hydrogen powers worker and verifier agents building and testing in parallel workspaces. On Terminal-Bench 4.0, Google says Orchestrator reached near-Carbon task success while using 24% less quota on average and finishing 19% faster. Google also reports safety gains for agentic work, including large reductions in overreach and in misaligned actions in its internal evaluations.
What changes in Search
Google calls Hydrogen its biggest model improvement to Search yet. In AI Mode it is the default model, and Google says it better understands multi-part questions, highly specific requests and hard-to-read typos, returns more visual responses — images, comparison tables and custom tools — and widens the range of links surfaced through its query fan-out technique, which breaks one question into sub-queries. TechCrunch notes the practical pitch behind the launch: the most powerful model is not always the one you need, and a faster, cheaper, reasonably smart one often finishes the job sooner.
Why it matters
Most real AI usage is not frontier research — it is thousands of routine coding, research and office tasks where cost and latency decide what gets deployed. A Gemini 4 model at Flash-style pricing that Google claims edges out its own flagship on a general index, and that ships the same day into Search, the Gemini app and the API, puts direct pressure on every rival’s fast tier. The thing to watch is whether independent testing confirms Google’s benchmark claims once developers have had Hydrogen for a few weeks.
Frequently asked questions
How much does Gemini 4 Hydrogen cost?
$1.50 per million input tokens and $7.50 per million output tokens, priced flat across its full 1 million-token context window, according to Google.
Where can I use Gemini 4 Hydrogen?
Google says it is rolling out globally in Google Search for all users, in the Gemini app, across all tiers of AI Studio and the Gemini API, in Gemini for business, and in Antigravity for paid Google AI Pro and Ultra subscribers.
How is Hydrogen different from Gemini 4 Carbon and Argon?
Carbon and Argon are Google’s frontier models for deep, long-horizon reasoning. Hydrogen is built for rapid execution of everyday coding and agentic workflows, and the three can be combined — with Carbon coordinating and Hydrogen running worker tasks — through the new Orchestrator in Antigravity.
Sources: Google, TechCrunch, 9to5Google