Google has launched Gemini 4 Hydrogen, a fast model for everyday coding and agentic work, while Carbon and Argon remain focused on deeper, long-horizon reasoning. It is rolling out globally across every Google AI Studio and Gemini API tier, paid Android Studio, Gemini for business, the Gemini app and Search. Paid Google AI Pro and Ultra subscribers can also use it in Antigravity, with Gemini in Sheets access starting on those plans.
Hydrogen costs $1.50 per million input tokens and $7.50 per million output tokens, flat across its 1 million-token context window. Its low, medium, high and xhigh thinking levels let users balance quality, latency and cost. Google says Hydrogen scores 51 on the Artificial Analysis Intelligence Index at the default high setting, beating Argon's 50, and reaches 54 at xhigh.
Introducing Gemini 4 Hydrogen, an extremely capable and fast model built to bring Gemini 4 to everyone!
— Logan Kilpatrick (@OfficialLoganK) November 12, 2026
Available to all tiers of Google AI Studio, Gemini API, Gemini for business, the Gemini app, Google Search, and paid Google AI Pro + Ultra subscribers in Google Antigravity. pic.twitter.com/LwDPYi1UCq
Antigravity is also gaining Orchestrator, which switches between Carbon and Hydrogen within a session. Carbon coordinates parallel agents while Hydrogen powers worker and verifier agents in isolated workspaces. On Terminal-Bench 4.0, Google says this setup approached Carbon's task success with 24% less quota and 19% faster average completion.
We’re rolling out Orchestrator in @Antigravity to Google AI Ultra subscribers, letting you run multiple models in a single session instead of routing the entire task to one model.
— Google (@Google) November 12, 2026
Orchestrator pairs a frontier model with a fast execution model: Carbon sets the plan, hands off… pic.twitter.com/AR7ZHcx6Nc
Against Gemini 3.8 Flash, Google says Hydrogen is 2.5 times faster at high, scores 57% versus 18.2% on Terminal-Bench 4.1 and cuts fabricated claims by more than fivefold. It targets repository coding, debugging, search, deep research, data analysis and multimodal work, and is designed to recover from tool failures. Hydrogen can accept 1 million tokens and produce up to 1 million, scoring 53.4% on Multimodal Knowledge Agent and 71.6% on Agents' Last Exam v1.1.
Introducing Gemini 4 Hydrogen, cutting-edge intelligence for everyone! Now in @GeminiApp. It's fast. It's smart. It's a big jump. Enjoy! pic.twitter.com/5AAIN6ryyh
— Josh Woodward (@joshwoodward) November 12, 2026
In Gemini, Google says Hydrogen strengthens instruction following and tool use for writing, research, files and connected apps. Search users worldwide get more intuitive answers, sharper query understanding, visual responses and broader links through expanded query fan-out. Google also reports up to 81% less agentic overreach, up to 79% less under-service and a 96.7% honesty rate in its internal evaluation, plus greater resilience to indirect prompt injection.
Google is positioning Hydrogen as the rapid-execution layer across its developer, enterprise, productivity, consumer and search products, not a replacement for the frontier reasoning models. The rollout ties Google DeepMind's model work directly to the company's main AI development and consumer surfaces.
Sources and related context
- Gemini 4 Hydrogen and Orchestrator in Google Antigravity: Supports: Google Antigravity describes Orchestrator's dynamic Carbon and Hydrogen routing and access for paid subscribers.
- Google Search gets a major upgrade with Gemini 4 Hydrogen: Supports: Google details the global Search rollout, improved query understanding, visual responses and broader links.
