Google is shoving Gemini deeper into people’s lives with Spark, a personal AI agent for subscribers willing to cough up the cash.
According to the Wall Street Journal Google has begun rolling out Gemini Spark, a personal agent it says can navigate a user’s digital life and act on their behalf. The agent will work across many of Google’s products and run on the company’s cloud infrastructure.
Google parent Alphabet chief executive Sundar Pichai said: “We’ve been bringing agents to developers and enterprises for a while now. “We are super focused on bringing that frontier capability of agents safely and securely to consumers so that they work for everyone.”
Google has tested Spark with a limited number of users and plans to make it available to AI Ultra subscribers next week. That new subscription tier costs $100 a month, which isn’t exactly loose change to ask a bot to meddle with your calendar.
Spark was announced at Google’s annual developer conference on Tuesday, as AI outfits Anthropic and OpenAI chase agents that make choices without constant human poking.
Google has unveiled Gemini Omni, a tool that creates videos from text, audio and images.
Google DeepMind chief AI architect and chief technology officer, Koray Kavukcuoglu, said: “It can create videos that are very, very high level, and you can interact with that video afterwards. You can imagine that this builds a tutorial-like capability, very much when you are exploring something.”
High-quality AI video has long been a compute-guzzling pain, with tools needing enormous amounts of processing power. OpenAI earlier this year scrapped Sora, its video-generation tool, partly to free up more computing resources for other projects.
Shares in Alphabet have surged since the November launch of Gemini 3, a model that helped Google claw ahead in the AI bunfight. OpenAI’s ChatGPT remains the most popular AI chatbot, while Anthropic’s Claude is regarded as one of the best coding models.
Investors still see Google as a serious rival, mostly because it has cloud, chips, search and a mountain of cash.
Google aslo introduced Gemini 3.5 Flash on Tuesday, a model it says is suited for coding and agentic work. The company claims Gemini 3.5 Flash runs four times faster than other frontier models.
Gemini 3.5 Pro, a stronger but slower model, is due to roll out in the coming weeks.
Google showed off voice-driven smartglasses that summon Gemini for directions, translations, texts, photos and tasks such as ordering from DoorDash.
The move targets Meta, which sold more than seven million pairs of Ray-Ban AI glasses in 2025. It comes 11 years after Google binned the consumer version of Google Glass, which was widely mocked and deserved it.
