Samsung has introduced Vision AI Companion, a generative AI-powered evolution of Bixby for its 2025 TVs, delivering on-screen, real-time assistance to address user queries about current content and broader household needs.
The Vision AI Companion operates as a conversational layer integrated directly into Samsung’s latest television lineup, enabling viewers to interrogate what appears on screen without switching devices or interrupting playback. Samsung outlines use cases including identifying “what an actor is famous for,” attributing “who painted a particular artwork,” or retrieving “the final score of a football game” referenced or displayed during broadcasts. Responses appear visually on the TV, aligning information with the viewed material.
Samsung extends the assistant’s role beyond content recognition. Vision AI Companion provides structured recommendations for TV shows and films, informed by user requests and context from ongoing viewing. It also supplies cooking guidance, delivers travel tips tailored to stated destinations or interests, and supports discovery of local restaurants, consolidating tasks commonly spread across phones or computers into a single screen-based interface.
The system relies on generative AI models, named by Samsung as Microsoft Copilot and Perplexity, to support natural language understanding, contextual memory, and follow-up questions during extended interactions. Within the television software environment, Vision AI Companion is positioned as a central hub coordinating other Samsung AI capabilities, such as automated picture optimization and real-time translation features, which are accessible through conversational prompts.
First announced at the IFA trade show in September, Vision AI Companion is now rolling out across Samsung’s 2025 TV range with support for 10 languages, including English, Korean, and Spanish. With no smart speaker line, Samsung uses the TV—alongside connected appliances like fridges and washing machines—as its primary in-home interface.
