Gemini in Chrome Android India Auto Browse.
On phones, the assistant does what it already does on a laptop. It can summarise a long article, answer questions about the page you’re on and pull in Google apps like Calendar and Keep without switching tabs. Nano Banana image editing comes along too, so a picture on a web page can be reworked with a typed prompt. Hindi, Tamil, Bengali and several other Indian languages are supported. Auto browse shows its plan before it touches anything
The Android rollout follows the desktop launch in India earlier this year.
AI Pro subscribers get up to 20 auto browse tasks a day, while Ultra users get 200. There are daily caps. The feature does not work in Incognito mode.
Google calls the agent experimental and says its models are trained to spot threats such as prompt injection. Google has brought Gemini in Chrome to all Android users in India and switched on auto browse, the assistant’s agentic mode, for paying subscribers on desktop and Android. Auto browse is limited to Google AI Pro and AI Ultra plans and lets Gemini carry out multi-step tasks on websites by itself, such as booking parking for an event, filling out a long form or ordering the week’s groceries. Before a task begins, Gemini lays out a plan for the user to check, including which sites it intends to use and what personal details it may share with them. It starts only after the user clicks Start Task, and works in a separate tab marked with an auto browse icon. Users can take over at any point, and Gemini stops to ask for confirmation before some sensitive steps. With permission, it can also sign in to sites through Google Password Manager without seeing the passwords. Its support page is blunt about where the buck stops, though: users are responsible for what Gemini does during a task, unintended purchases included. You use AI every day. Now get your AI Quotient. Take the AIQ test.