Samsung brings generative AI Bixby to its TVs: What’s new and how it works

Announced earlier at IFA 2025 in Berlin, Samsung’s Vision AI Companion lets viewers ask questions about on screen content and more. Now rolling out across the 2025 TV line up, it introduces a conversational interface that understands natural, context aware queries about what is on the screen and related topics. Built on a generative AI upgrade to Bixby, the assistant returns visual answers directly on the TV, reducing the need to reach for a phone or navigate nested menus.

At its core, Vision AI Companion is designed to understand what you are watching and provide timely information. Viewers can ask who an actor is, what a painting represents, or the final score in a live match. The assistant can also move beyond the current programme with recommendations for films and series, simple recipe ideas, travel tips, and local restaurant suggestions. Responses appear with visual elements on the television, and the system supports follow up questions that preserve context within the same conversation.

According to Samsung, the feature blends multiple large language models, including Microsoft Copilot and Perplexity, to generate contextual answers. It launches with support for ten languages, including English, Korean, and Spanish. Vision AI Companion works across Live TV, Samsung TV Plus, and major streaming applications, with the goal of maintaining a consistent experience regardless of the content source.

Samsung is positioning Vision AI Companion as a hub for its wider set of TV intelligence features. Live Translate offers real time dialogue translation for supported content. AI Gaming Mode adjusts picture and sound during gameplay. Generative Wallpaper adapts ambient visuals to user preferences. Image and audio enhancements such as AI Picture, AVA Pro, and AI Upscaling Pro continue to handle clarity, dialogue emphasis, and resolution improvements across varied inputs.

Availability spans the 2025 line up, including Neo QLED, Micro RGB, OLED, QLED step up models, Smart Monitors, and The Movingstyle. All devices run on One UI Tizen. Samsung states that supported televisions will receive up to seven years of operating system updates covering features and security, which is notable given the long ownership cycles for large screen devices.

The approach places a consolidated voice and visual assistant on the largest screen in the home. Televisions have offered voice search and basic controls for years, but the generative model aims to broaden the scope of interactions and keep them within the viewing context. The visual presentation of answers on the TV is intended to minimise disruption, so viewers can get quick context and carry on watching.

Adoption will hinge on day to day reliability. Key factors include how accurately the assistant recognises on screen elements, how well it maintains conversational context across third party apps, and how effectively it handles regional content. Initial language coverage is a start, and broader support will matter for reach in more markets. If performance proves consistent and responses remain timely and clear, Vision AI Companion could shape how viewers search, learn, and decide what to watch without leaving the sofa.