900 million monthly users, a new design, a model for generating videos, and a personal AI agent that writes emails, analyzes invoices, and doesn’t need your presence. Google has shown where Gemini is headed.
900 million monthly users, a new design, a model for generating videos, and a personal AI agent that writes emails, analyzes invoices, and doesn’t need your presence. Google has shown where Gemini is headed.
Google has launched Gemini 3.5 Flash, the first of a new generation of Flash that combines high quality with speed. At the same time, the company has completely redesigned the application interface: the new design is called Neural Expressive and includes dynamic animations, new typography and haptic feedback.
Key interaction change: You can now seamlessly switch between text chat and live voice conversation. The microphone has been redesigned to not interrupt your thoughts in mid-sentence. Regional dialects will be available soon, allowing you to choose a voice that is close to your own.
Instead of solid text, Gemini now generates responses dynamically — with images, interactive timelines, videos, and graphs. Neural Expressive is now available globally on the web, Android, and iOS.
Gemini Omni is a new model for video generation. It takes text, images, and video clips as input and transforms them into cinematic content. You can upload your own footage, add templates, and edit the result through simple text queries — without editing programs or technical knowledge.
One feature is the creation of a personal AI avatar that looks and sounds like you. Gemini Omni is now available to Google AI Plus, Pro, and Ultra subscribers worldwide.
Daily Brief is an agent that prepares a personalized digest every morning. It scans Gmail, Calendar, and connected apps, prioritizes messages based on your goals, and suggests specific next steps. Not just a summary — it’s a plan for the day.
The user can train the agent through reactions — likes and dislikes — and it gradually adapts. Daily Brief is currently available to Google AI Plus, Pro, and Ultra subscribers in the US.
The main announcement of the conference is Gemini Spark. It is a cloud-based AI agent that continues to perform tasks even when you close your laptop or lock your phone. It is deeply integrated with Gmail, Docs, Slides, and other Workspace tools.
What Spark can do: automatically analyze credit card statements and find hidden subscriptions; monitor school letters and send you and your partner a daily summary with deadlines; collect meeting notes, format them in a Google Doc, and send a project start letter. That is, it can execute complete work scripts, not individual requests.
Spark will also get new MCP integrations — today, connections to Canva, OpenTable, and Instacart are launching. In the near future, the agent will learn to send SMS and emails on its own, as well as control the browser.
Spark operates solely under your control: you decide which applications to connect, and it asks for permission before any important action — for example, before sending an email or making a payment.
Spark will begin testing with trusted users this week, and will become a beta for Ultra subscribers in the U.S. next week. A macOS version with local file support will follow in the summer.
