Google has started releasing new AI tools for Gemini that let the assistant analyze your phone screen or camera feed in real time. These features, available to Google One AI Premium subscribers, allow Gemini to instantly answer questions about what it sees. A Google spokesperson confirmed the rollout, which follows the company’s “Project Astra” demo from last year.
The update makes Gemini more interactive, turning it into a visual helper. For example, you can point your camera at an object, and Gemini will describe it or suggest ideas. Like Reddit members with Xiaomi phones, early users have already tested the screen-reading tool. Google plans to expand access to more users soon.
How Gemini’s New Video Tools Work
Gemini’s screen analysis feature lets the AI scan whatever is on your phone display. If you are reading an article, Gemini can summarize it. If you are browsing products, it can compare prices. The live video tool uses your camera to analyze the real world. In a demo video, someone asks Gemini for advice on picking a paint color for pottery. The AI suggests options based on the camera feed.

These tools are powered by Project Astra, Google’s effort to create a universal AI assistant. Unlike basic voice helpers, Astra processes visual data to understand context. This means Gemini can handle complex tasks, like explaining a graph in a presentation or identifying plants during a walk.
Why Google Is Ahead of Amazon and Apple
According to the company, the data related to video processing remains private because Google operates these functions directly from the device. Previous AI error cases have criticized the company, yet the organization guarantees more vigorous system verification. Users remain concerned about their camera devices being constantly turned on.
Google One AI Premium subscribers using the tools must have a $20 monthly subscription. The premium device requirement may dissatisfy users who do not subscribe to Google One AI Premium, but Google might introduce alternative plans later. Currently, the emphasis remains on verifying the system with paying customers.
Privacy and Availability Concerns
Google says all video processing happens on-device, so your data stays private. The company has faced criticism over AI errors, but it promises stricter checks now. Still, users might worry about cameras always being active.
The tools are only for Google One AI Premium subscribers, who pay $20 monthly. This could frustrate Gemini’s users, but Google might offer cheaper plans later. For now, the focus is on testing with paying customers.

What This Means for the Future of AI Assistants
Gemini signifies how artificial intelligence technology transitions beyond voice functionalities into visible partnership capabilities. Additional software updates might enable the device to operate smart home appliances using a camera function and potentially execute cooking steps automatically based on visual observation. Google’s extensive initial market presence may influence future competitor AI development towards increasing visible features in their products.
Different users of Gemini’s video features will help the company enhance its tools based on received feedback. Such successful development marks the start of artificial intelligence systems that genuinely perceive and understand environmental surroundings.