Key Points:
- Google is integrating Gemini AI into Google TV to enhance search and discovery, offering personalized recommendations based on user data and natural-language queries.
- Gemini allows users to ask conversational questions through far-field microphones, turning TVs into interactive smart hubs.
- Google says the goal of Gemini on TV is to drive scale and prove value before determining its long-term monetization strategy.
Google is readying plans to push out its artificial intelligence assistant Gemini to most smart TVs and streaming devices running Android with the Google TV user interface between now and next year, an executive confirmed to The Desk.
In an interview, Google’s Vice President & General Manager of Android TV Shalini Govil-Pai said some newer-model Android TV devices will receive an update that brings Gemini capabilities to the TV by the end of this year, with most Google TV devices receiving the same update throughout 2026.
Govil-Pai says Android TV will continue to support the Google Assistant for remedial tasks, like turning on smart light bulbs, while Gemini will handle more-complex queries, like trying to find new things to watch or surfacing a particular movie or TV show based on a vague description of the actor or plot.
Gemini is coming to a handful of smart TVs and streaming devices this year, including newer-model TV sets made by TCL and Hisense along with standalone streaming hardware like the Google TV Streamer and Walmart’s Onn TV 4K Pro, the latter of which costs around $50. Some of those devices are already in homes and will receive Gemini through a software update if it isn’t already available.
Other devices will receive Gemini throughout the next year, Govil-Pai affirmed. The integration isn’t entirely dependent on hardware requirements, because Gemini queries are performed in the cloud, and then returned to users over the Internet.
“It’s all done in the cloud — we don’t do it locally, so the query goes to the cloud, and then we decide whether it is answered by Google Assistant or Gemini,” she said.
By utilizing the cloud to return queries, Gemini features can be integrated into low-cost and older devices that currently run Google TV. Some extremely low-power devices won’t be able to receive the integration, Govil-Pai said, but the number of incompatible devices is expected to be very low.
The integration of Gemini into Google TV sets and streamers comes at a time when major tech companies are leaning into artificial intelligence as a way to solve numerous issues that have plagued the streaming industry, including search and discovery.
Govil-Pai said Gemini can cut through the overwhelming amount of content on streaming platforms today — including Google’s own, like YouTube and Google TV Freeplay — by giving people recommendations based on limited descriptions coupled with their watch history and other data.
Govil-Pai used the hypothetical situation of a person searching for a movie where the main character, played by a handsome actor, races cars. With Google Assistant, a list of possible movies and the apps where they are available appears — but Gemini should know, based on various signals, whether a person intended to search for “F1: The Movie” starring Brad Pitt or “The Fast and the Furious” with Vin Diesel.
Gemini’s integration goes beyond content recommendations: Govil-Pai envisions a world where people interact with their TVs the way they current search for things on Google or ask questions to their smart speakers. Newer-model TV sets with Gemini integration feature far-field microphones that give users the ability to leave their TV sets on and bark questions and commands at it throughout the day — just as they would a smart hub.
“What we’re finding is that people are gravitating toward a new pillar of queries, and that’s knowledge-based,” Govil-Pai said. “We saw a recent query, can dogs eat broccoli? In the past, you’d never think to ask your TV about that — but, now, you can do it right in your living room, and have natural responses delivered.”
In other words, Gemini won’t simply deliver a bunch of content tiles to YouTube videos — it will actually answer a question posed, using conversational language, based on a variety of sources, the way Gemini and other large language model-based chatbots work on the web.
People have already shown a level of comfort looking for things via chatbots, which have exploded in popularity over the past few years. Gemini has around 350 million monthly active users, according to documents Google submitted with a civil court in April. By comparison, the number of people using Gemini was around 9 million users just six months prior, TechCrunch reported.
Gemini and its competitors like OpenAI’s ChatGPT and Anthropic’s Perplexity are changing the way people search for things with their computers and their phones. Google doesn’t see why the same shouldn’t be true for smart TVs and streaming devices, since the technology can be adapted to fit most uses cases there as well.
When asked how Google intends to make a buck off Gemini’s integration into Google TV, Govil-Pai said the company will evaluate how people use Gemini on their smart TVs and streamers before deciding how it can accomplish certain business goals.
“The goal for our platform is not necessarily to drive revenue, but to drive distribution and scale,” she said. “When it comes to the business model, we first want to prove that there’s value for Gemini, and then we will figure out the right business model.”
—
Read more:
