Google Launches "Real-Time Video" in Gemini AI

SEAToday.com, Jakarta - Google has expanded the capabilities of its generative artificial intelligence (AI) service, Gemini, by introducing a real-time video feature.

This new functionality allows users to engage with AI through their smartphone cameras, enabling real-time interactions based on visual input.

As reported by The Verge, the feature enables Gemini to process what a smartphone camera captures and provide immediate responses related to the visual content. This advancement aligns with Google's broader efforts to enhance AI-powered assistance through more dynamic and interactive user experiences.

Google spokesperson Alex Joseph confirmed the launch, highlighting that the feature builds on the company’s previous research under Project Astra. The project, which began nearly a year ago, explores ways to improve AI's ability to interpret and respond to real-world visual information.

The feature was first discovered by a Xiaomi smartphone user who accessed Gemini’s new capability and shared their findings on the Reddit community. Other users soon followed, demonstrating Gemini’s ability to read text directly from their phone screens and provide contextual responses.

At the Mobile World Congress (MWC) in March 2025, Google had already announced plans to introduce this capability to select Gemini Advanced users. The real-time video interaction is part of a broader push to make AI assistants more intuitive and versatile.

With this update, users can now point their smartphone cameras at an object, document, or environment and receive AI-driven insights in real time. Google has also released a demonstration video showcasing the feature’s practical applications, emphasizing its potential in assisting with everyday tasks.

The introduction of real-time video analysis places Google ahead of its competitors in the AI assistant space. While Amazon is preparing an early-access launch for Alexa Plus, and Apple works on an upgraded version of Siri, Google’s Gemini has already begun integrating advanced visual processing into its AI ecosystem.

Writer: Andi Raisa Malaha Thambas

1 SELENGKAPNYA