Google has launched Gemini 4 Hydrogen, moving away from its previous Flash branding to introduce a standalone upgrade designed for speed and broader access. The model offers a native 1 million-token input window along with up to 1 million output tokens. It is available globally in Google Search for all users, where it serves as the default engine for AI Mode and handles challenging queries in AI Overviews. For developers, access costs $1.50 per million input tokens and $7.50 per million output tokens. [5][7]
The new architecture aims to better parse complex, multi-part questions, typos, and specific user instructions. Google claims Hydrogen cuts fabricated claims by more than 5x and finishes tasks 2.5 times faster at its high default setting than Gemini 3.8 Flash. In Search, the model runs simultaneous subtopic queries to surface a wider range of web links, paired with visual elements like comparison tables and images across supported languages. [5][7]
Beyond Search, Hydrogen is categorized under everyday help in the Gemini app, where free users can access it through automatic model selection. In Google Antigravity, paid Google AI Pro and Ultra subscribers can use Hydrogen alongside an Extra thinking tier. It also powers a feature called Orchestrator, which allows the more advanced Carbon model to offload routine steps to Hydrogen to reduce overall token expenses. [5]