Editorial illustration for Google DeepMind adds Gemini-powered cursor to Chrome for visual queries
Google DeepMind adds Gemini-powered cursor to Chrome for...
For months, the AI industry has been obsessed with better prompts. Google DeepMind just scrapped the whole premise. Starting today, in Chrome, you don’t type your question, you point at it.
A Gemini-powered cursor reads the visual and semantic context of whatever you’re hovering over, then acts on it without a single word of instruction. Select a few products on a page, and the AI compares them. Point to an empty corner of your living room photo, and it visualizes a new couch.
This isn’t a faster search bar. It’s a fundamental inversion: instead of dragging context into an AI window, the AI follows your pointer across every app you already use. The system is built on four principles, among them, “Turn pixels into actionable entities.” That means the cursor doesn’t just see images; it identifies objects, dates, places, and relationships, turning static screen content into live, queryable elements.
Two demos are live in Google AI Studio today. Gemini in Chrome rolls out now. Later this year, Magic Pointer arrives for Googlebook.
The era of the prompt is ending. The era of the pointer has begun.
Starting now, instead of writing a complex prompt, users can use their pointer to ask Gemini in Chrome about the part of the webpage they care about. For example, selecting a few products on a page and asking to compare them, or pointing to where they want to visualize a new couch in their living room.
Key Takeaways
- Google DeepMind introduces experimental demos of an AI-enabled mouse pointer powered by Gemini that captures visual and semantic context around the cursor -- no manual prompting required.
- The system is built on four principles: Maintain the flow, Show and tell, Embrace the power of “This” and “That”, and Turn pixels into actionable entities.
- “Turn pixels into actionable entities” is the key technical idea -- the pointer converts on-screen content into structured entities like places, dates, and objects that users can act on instantly.
- Two live demos are available now in Google AI Studio (image editing and map search); Gemini in Chrome is rolling out today, with Magic Pointer for Googlebook coming later this year.
- The core design shift: instead of users dragging context into an AI window, the AI follows the cursor across every app the user is already working in.
Check out the Technical details.
The cursor has always been a tool of intention, point, click, drag, done. Now it becomes something else: an extension of thought. By weaving Gemini directly into the pointer’s path, DeepMind collapses the distance between seeing and acting.
You no longer translate curiosity into a prompt; you just look, and the AI follows your gaze. That shift is quiet but profound. It turns every pixel into a question waiting to be answered, every selection into a conversation.
The demos are live. The rollouts have begun. What happens next is not about faster searches or smarter assistants.
It is about remaking the interface itself, from a place where we ask for help to a place where help finds us, exactly where we already are.
Related Reading
Browse all Google news →
Research & Benchmarks Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks Databricks finds multi-step agents beat single-turn RAG by 21% to 38% on STaRK
Research & Benchmarks Nvidia's DLSS 4.5 beta adds 6x Multi Frame Generation for RTX 50 GPUs
Related Topic NVIDIA and Google Cloud let developers scale AI from prototype to production
Market Trends
Related Topic Google tests visual 'magazine-style' UI for Gemini 3 Pro users
LLMs & Generative AI
From Archives BaLoRA adds Bayesian uncertainty to low‑rank adaptation, but lags fine‑tuning
From Archives Community review tools guide novices in AI research, study finds
From Archives Android adds Gemini AI agents for trip booking, shopping lists, summarizing
Business & Startups
From Archives Google stops attack after AI finds zero‑day; China, North Korea also using AI
Open Source
Common Questions Answered
How does the Gemini-powered cursor in Chrome work without requiring typed prompts?
The Gemini-powered cursor reads the visual and semantic context of whatever you hover over on a webpage, then performs actions based on that understanding without needing any typed instructions. This eliminates the traditional need to formulate and type out prompts, allowing users to simply point at content to trigger AI responses.
What are some specific use cases for the new Gemini cursor feature?
Users can select multiple products on a page and have the AI compare them automatically, or point to an empty area in a photo of their living room and have the AI visualize how a new piece of furniture would look in that space. These examples demonstrate how the cursor enables both analytical tasks and creative visualization without explicit instructions.
How does Google DeepMind's approach differ from the industry's focus on better prompts?
Rather than continuing to emphasize crafting better text prompts, Google DeepMind has fundamentally changed the interaction model by integrating Gemini directly into the cursor's functionality. This shift allows users to act on their curiosity through visual pointing and selection rather than translating their thoughts into written prompts.
What is the philosophical significance of turning the cursor into an extension of thought?
By collapsing the distance between seeing and acting, the Gemini cursor transforms how users interact with digital content, turning every pixel into a potential question and every selection into a conversation. This represents a shift from explicit instruction-based interaction to intuitive, gaze-following AI assistance that mirrors natural human thought processes.
Further Reading
- Reimagining the mouse pointer for the AI era — Google DeepMind
- Chrome gets new Gemini 3 features, including auto browse — Google Blog
- Computer Use - generateContent API — Google AI for Developers
- AI in Chrome: help right where you need it — Google Chrome
- Artificial Intelligence in Chrome — Chrome Developer
