Google introduces new AI features for Gemini Live with screen sharing capabilities: Report

After its initial announcement in this month’s MWC 2025, Google has now reportedly started introducing real-time AI features to Gemini Live, which lets you screen or through your smartphone camera and answer questions about either in real-time. Developed by Google DeepMind under Project Astra, these features come with live multimodal data processing capabilities and introduce a new way for users to engage with Gemini.

Google’s new Gemini Project Astra features: Here’s how it works

Google has started introducing real-time AI features to Gemini Live, enabling the AI to interpret screens and smartphone camera feeds instantly. Google’s spokesperson, Alex Joseph, said to The Verge that these capabilities are now accessible to some Google One AI Premium subscribers. Moreover, it builds on Google’s earlier Project Astra initiative, which was first showcased nearly a year ago at the Google I/O 2024.

Additionally, a Reddit user, Kien_PS, recently reported spotting this feature on their Xiaomi device, as noted by 9to5Google. Later, the user uploaded a video demonstrating Gemini’s new ability to read on-screen content. Another major feature now being introduced is live video interpretation. With this, Gemini can analyze a smartphone’s camera feed in real-time and provide relevant insights.