Summary: On January 30, 2026, Google shipped Auto Browse in Gemini in Chrome as part of its monthly Gemini Drop, letting an agent handle multi-step browser tasks while the user retains control. Android Central reported access for Google AI Pro and AI Ultra subscribers. The article frames the browser as an existing agent entry point with login state, payment information, and task context already attached.
Google kept its monthly appointment on January 30, 2026, shipping the January “Gemini Drop” with Auto Browse in Gemini in Chrome as the centerpiece. The company’s own pitch: “tell Gemini what you need and watch it handle the rest for you” — multi-step jobs that live in the browser, from booking appointments to planning parties, can now be handed wholesale to an agent. Android Central came away impressed, suggesting it might be “Gemini’s biggest month of updates yet.”
What deserves attention here is not the feature itself but where it was put. The browser is the doorway people keep open all day, so seating an agent inside Chrome moves agentic AI out of the chat window and into workflows people already run. Set against the option of opening yet another app to ask a question, this route ends closest to the task actually being finished.
Auto Browse: An Agent Takes Over the Browser
The demos Google chose are appointment booking and party planning, and the company is careful to note that the user keeps control the whole way through. Where Auto Browse departs from a typical “AI feature” is the unit it works on: not a query but a task — many steps long, crossing sites, and carrying state as it goes. You name the goal; the clicking, filling, and confirming belong to the agent. This is the 2026 agent race moving out of chat interfaces and into established entry points such as browsers and operating systems, made concrete — the very trend our 2026 opening outlook called out.
Eligibility: Subscription Tiers and Platform Support
Per Android Central’s roundup, Auto Browse is reserved for Google AI Pro and AI Ultra subscribers. The Drop’s other items sit behind different thresholds: the SAT practice tests cost nothing, while Gemini in Chrome itself — running on Gemini 3 — is open to users on macOS, Windows, and Chromebook Plus, with a collapsible side panel that hooks into Connected Apps such as Gmail, Calendar, and Maps. Android Central adds that January also saw the cheaper Google AI Plus tier reach the United States for the first time, completing a three-tier paid lineup.
Also in the Drop: Veo 3.1, SAT Practice, and Shopping
Everything else in the January Drop:
- Veo 3.1: turns reference images into moving video inside the Gemini app, with stronger dialogue and narrative output
- SAT practice tests: free full-length exams available on demand, built with content vetted by The Princeton Review and paired with real-time personalized feedback; “SAT only for now,” though more tests are planned
- Chrome shopping upgrades: researches products against the tabs you currently have open and suggests complementary items
- Personal Intelligence (beta) rounds out the month’s list: it links data across your Google apps to offer tailored assistance, which apps to connect is the user’s choice, and Google says personal data stays out of model training. The feature debuted on its own in mid-January; this Drop merely folds it into the monthly cadence
What the Monthly Rhythm Buys
Gemini Drop is Google’s chosen answer to continuous delivery for AI products: one predictable bundle each month in place of a handful of big launches a year. Users come to expect new capability as routine rather than occasion. Inside Google, the format holds every team to a monthly settling of accounts. At a moment when Microsoft, OpenAI, and startups are all contesting the agent entry point, a shipping rhythm you can count on is itself a competitive edge — agent products are the textbook case of “accuracy comes from usage,” and iteration speed decides how deep the moat goes.
Takeaways for Developers and Product Teams
Three things stand out. One: Auto Browse lifts the UX bar for task-level agents to first-party height, and browser-automation products (form filling, workflow bots, data extraction) will now meet Google’s own agent head on — standing apart means going deeper, into specific workflows and enterprise integrations. Two: the tiering shows Google books agentic capability as paywalled content — Auto Browse belongs to Pro and above while the free tier receives acquisition tools in the form of education and video — a “free drives scenarios, paid sells agents” split worth copying in pricing strategy. Three: never underestimate the browser as an agent container — it comes with login state, payment info, and context already attached, which erases the hardest stretch of agent cold start.