CES 2026: Gemini for Google TV Gets Upgraded With Nano Banana’s Image Editing Feature

Gemini for Google TV is getting a major upgrade. The artificial intelligence (AI) assistant for the streaming platform was introduced in September 2025 to replace the older Google Assistant and has been steadily expanded to more smart TVs. Now, at the Consumer Electronics Show (CES) 2026, the Mountain View-based tech giant is previewing a set of new features that will soon be rolled out to users. Among them is the addition of the AI-powered image model Nano Banana and the video generation model Veo.

Gemini on Google TV Gets a Major Upgrade

In a blog post, the tech giant announced the new features Google TV users can expect to see on their compatible smart TVs in the coming months. The company said the Gemini capabilities will first be rolled out to select TCL devices, and then to other TVs over the coming months. Since Gemini's expansion on Google TV is not complete, these are still available in select regions and support select languages. Additionally, Android TV OS 14 or newer is required for the AI features.

There are essentially four new features that users will soon see. The first is the new visually rich framework introduced with the Gemini 3 Pro AI model, which generates images, slideshows, and interactive visual elements when responding to users. Users can also use the Deep Dive option to gain an in-depth understanding of a complex topic.

Gemini on Google TV can now search users' Photos library to surface specific people or moments. Users can find desired images via natural language voice-based prompts. Gemini can also access Photos Remix and create a slideshow using the user's memories.

These photos can also be edited and reimagined using Nano Banana and Veo models, which are now coming to Gemini on Google TV. Alternatively, these models can also be used to generate original creations directly on the TV.

Finally, users can have a natural language voice conversation with Gemini to change and optimise the settings based on what they are watching. For instance, if someone tells Gemini “the screen is too dim” or “the dialogue is lost,” it will understand the context and adjust the picture and volume.