Google has announced Gemini 4 Hydrogen, its latest AI model. Adapted from Gemini 4 Carbon and focused on speed, Google claims that Hydrogen also surpasses its frontier Gemini 4 Argon model when using its default thinking level.
“While our frontier models, Gemini 4 Carbon and Gemini 4 Argon, excel at deep, long-horizon reasoning on your hardest problems, Hydrogen is built for rapid execution across diverse agentic workflows,” wrote Koray Kavukcuoglu, senior vice president at Google DeepMind, in Thursday’s announcement.
Specifically, Hydrogen is intended to quickly conduct tasks such as writing code, running tests, and doing everyday work, as opposed to designing systems and fixing complex bugs. Set to a “high” thinking level by default, Hydrogen can also be adjusted to low, medium, and a new xhigh tier, which vary in quality, latency, and cost.
“Hydrogen is built to stand on its own as a model for everyday work — or to pair with Carbon and Argon as the high-efficiency execution engine in multi-model workflows,” wrote Kavukcuoglu.
The best tech deal this week
Products available for purchase through affiliate links. If you buy something through links on our site, Mashable may earn an affiliate commission.
Mashable Light Speed
Gemini 4 Hydrogen is rolling out globally from today, arriving in Google Search for all users. It will also be available for selection in the Gemini app’s model dropdown menu.
Google Search’s AI features were previously powered by Gemini 3.8 Flash, which launched in September. In comparison, Kavukcuoglu stated that Hydrogen will respond to queries with more visual responses, as well as a wider variety of relevant links. Hydrogen is also designed to better understand conversational prompts and follow users’ trains of thought, including understanding typos.
A comparison between Gemini 4 Hydrogen and Gemini 3.5 Flash, a previous AI model released in May.
Credit: Google
“Gemini 4 Hydrogen makes it easier to get things done, staying on track and following instructions more closely across multi-turn conversations,” said Kavukcuoglu. “In the Gemini app, Hydrogen makes your everyday assistant even more helpful — it’s better at following your guidance, calling the right tools, and getting you to a reliable output with less back-and-forth.”
Compared to 3.8 Flash, Google claims that Hydrogen is two times less likely to circumvent permission boundaries. The company further stated that it’s continuing to work on the AI model’s safety and security aspects such as jailbreak resistance, though noted that Hydrogen is still prone to hallucinations, like all generative AI agents.
