Google Gemini 4 Hydrogen and Gemini 4 Carbon bring fast everyday execution and deep frontier reasoning to Omnifact via Google Cloud in the EU and the Google API.

Google Gemini 4 Hydrogen and Gemini 4 Carbon bring fast everyday execution and deep frontier reasoning to Omnifact via Google Cloud in the EU and the Google API.

Feature Drop: Google Gemini 4 Hydrogen and Carbon Now Available

Published on November 16th, 2026

Gemini 4 Hydrogen and Gemini 4 Carbon are now available on Omnifact. Together, they expand Google's Gemini 4 lineup in two directions: Hydrogen brings fast Gemini 4 performance to everyday coding, writing, and agent workflows, while Carbon pushes reasoning depth further for complex engineering, scientific work, and multi-step analysis. Both models run with European data residency via Google Cloud in the EU and through the Google API.

Two New Roles in the Gemini 4 Lineup

  • Gemini 4 Hydrogen (Standard tier): Built for the work teams do many times a day, from answering questions across a Knowledge Base and drafting reports to writing code and running multi-step agent tasks. Hydrogen takes over from Gemini 3.8 Flash (Google) and Gemini 3.5 Flash (Google EU) as the default Google model for everyday work.
  • Gemini 4 Carbon (Premium tier): Google's strongest reasoning model, built for assignments where deep verification matters more than turnaround speed. Carbon spends more thinking tokens per turn to explore hypotheses, check edge cases, and validate fixes across repositories, spreadsheets, and dense documents.
  • Gemini 4 Argon (Premium tier): Introduced in October, Gemini 4 Argon stays in the lineup beside Carbon at the same price. Use Argon for fast, dependable complex work, and switch to Carbon when a problem pushes reasoning and verification to the limit.

Comparing the Gemini 4 Models

AreaGemini 4 HydrogenGemini 4 ArgonGemini 4 Carbon
Primary RoleFast everyday work, coding, and agent tasksEfficient frontier reasoning and complex workDeepest reasoning, verification, and complex engineering
Context Window1M tokens1M tokens1M tokens
Reasoning EffortLow, Medium, High, Extra HighLow, Medium, HighLow, Medium, High
Model TierStandardPremiumPremium

Adjustable Reasoning and Long Context

Both models include a 1-million-token context window and configurable Reasoning controls directly in the Chat Input model picker.

  • Extra High reasoning on Hydrogen: Gemini 4 Hydrogen adds an Extra High reasoning level alongside Low, Medium, and High. Keep it on Low or Medium for quick lookups and short drafts, or switch to High or Extra High when an agent task spans multiple files or sources.
  • Deeper verification on Carbon: Gemini 4 Carbon supports Low, Medium, and High reasoning levels and puts additional thinking depth toward root-cause debugging, architectural design, scientific reasoning, and thorough document checks.
  • Full-project context: With 1 million tokens of unified context, either model can work through a long contract, a set of meeting transcripts, or a Space synced via Connected Sources in a single conversation.

Available Today with EU Data Residency

Gemini 4 Hydrogen and Gemini 4 Carbon are available today for all Omnifact organizations via Google Cloud in the EU and the Google API. Through Google Cloud, requests are processed in the EU under zero data retention commitments, and every interaction in Omnifact remains protected by our Privacy Filter.

New organizations receive Gemini 4 Hydrogen, Gemini 4 Argon, and Gemini 4 Carbon in their default Google and Google (EU) setup. Existing team administrators can enable both models in Team Settings > Models or assign access by department using Group Management, keeping daily workflows on Hydrogen while reserving Carbon for teams that need maximum reasoning depth. Select either model in the Chat Input model picker to try it on your next task.

Share this article