Google rolls out Gemini 4 Hydrogen across Search and its AI tools

Google introduced Gemini 4 Hydrogen on November 12 and says the model is rolling out globally across Search, the Gemini app, developer tools and eligible business products. The practical change for most readers is in Search: Google says its upgraded AI experience can handle more complex, multipart questions and present structured answers with images, tables and links to relevant sites.

The company describes Hydrogen as a model built for rapid everyday and agent-based work, distinct from its more resource-intensive Gemini 4 Carbon and Argon models. Those are Google’s product-positioning claims, not independent proof that a given answer will be correct. As with other generative AI tools, important facts still merit checking against the linked original sources.

Google says the Search upgrade is available globally to all users. In the Gemini app, access is more nuanced: free users may encounter Hydrogen through automatic model selection when the app chooses it for a prompt, rather than necessarily selecting it directly. The company lists broader access for developers through the Gemini API and AI Studio, and for eligible enterprise and paid subscribers across its other products. Rollout language means an individual account or region may not display every feature at once.

For developers, Google lists a price of $1.50 per million input tokens and $7.50 per million output tokens across a context window of up to one million tokens. Those API charges are not a consumer subscription price. Google also released a model card describing the model’s inputs, outputs and limitations.

The announcement matters because it puts a newly released model into a mass-market search product immediately, while also offering it to software builders. The headline change is availability, not a benchmark contest: Google’s own performance comparisons should be treated as vendor claims until independently tested in real-world tasks.