Google adds Gemini 4 Hydrogen model to Search for faster, more nuanced answers
Google announced that its new Gemini 4 Hydrogen model will power Search, offering quicker responses, better handling of complex queries and richer visual results.
Google said its Search service will be upgraded with the Gemini 4 Hydrogen model, which it describes as a faster, lower-cost alternative to its larger frontier models. Hydrogen is built for “agentic workflows,” allowing the AI to carry out multiple steps autonomously. The company claims the model completes tasks 2.5 times faster than the previous Gemini 3.8 Flash model and outperformed the more powerful Gemini 4 Argon on a broad intelligence benchmark at its default high thinking level.
In Search, Hydrogen is intended to improve understanding of user intent, handle multi-part or highly specific questions, and correct confusing typos. The model’s reasoning capabilities also enhance the “fan-out” technique that breaks a single prompt into subqueries, helping users discover additional relevant sites. Visual responses such as images, structured data, comparison tables and custom tools are expected to appear more frequently in AI Mode.
Google noted that Hydrogen can operate independently for tasks like writing or testing code, or be combined with the Carbon and Argon models in more complex workflows. The model offers four thinking levels—low, medium, high (default) and an extra-high tier—designed to finish tasks faster with fewer output tokens. In multi-step tasks such as deep research and data analysis, Hydrogen ranked fifth on Agent Arena’s benchmarks.
The update will be rolled out globally to all Search users and will also be available through Google’s AI Studio, Gemini API, Gemini App and the paid AI Pro and Ultra tiers of the Antigravity development platform.