Google I/O 2025 Everything Announced So Far: Gemini Ultra, Veo 3, Imagen 4, AI Updates, More

Google's annual developer conference, Google I/O 2025, is underway at the Shoreline Amphitheatre in Mountain View. The event spans two days, showcasing Google's latest product announcements. Key highlights include updates on Android, Chrome, Google Search, YouTube, and Google's AI chatbot, Gemini. A separate event called The Android Show was also held to introduce new Android features.

Gemini Ultra is a new offering available only in the U.S. for now. It provides top-tier access to Google's AI-powered apps and services for $249.99 monthly. This package includes Veo 3 video generator, Flow video editing app, and the upcoming Gemini 2.5 Pro Deep Think mode. AI Ultra subscribers also benefit from enhanced limits on Google's NotebookLM platform and Whisk image remixing app.

AI Innovations and Tools

Deep Think mode in Gemini 2.5 Pro enhances reasoning by considering multiple answers before responding. Although details are sparse, it might resemble OpenAI's models that synthesise optimal solutions. Currently available to "trusted testers," Google is conducting safety evaluations before a broader release.

Veo 3, an AI model for generating videos with sound effects and dialogue, is now accessible through Google's Gemini chatbot app for AI Ultra subscribers. It surpasses its predecessor in footage quality and can be prompted with text or images.

Image Generation and Video Tools

Imagen 4 is another AI innovation from Google, boasting faster rendering than Imagen 3 and capable of creating detailed images in various styles and resolutions up to 2K. Both Veo 3 and Imagen 4 will support Flow, an AI-powered video tool aimed at filmmakers.

Google has announced that its Gemini apps have over 400 million monthly active users. New features like camera and screen-sharing capabilities are rolling out this week on iOS and Android via Project Astra.

Enhanced User Experience

Gemini Live will soon integrate more deeply with other Google apps such as Maps, Calendar, and Tasks. Users can upload private PDFs and images to Deep Research for comprehensive reports.

Stitch is an AI tool designed to help create web and mobile app front ends by generating UI elements and code from simple prompts or images. While it offers customisation options, it's somewhat limited compared to other coding products.

Developer Tools and Features

Jules assists developers in fixing code bugs by understanding complex code structures and managing tasks on platforms like GitHub. Project Mariner allows users to perform tasks online without visiting third-party websites by interacting with Google's AI agent.

The low latency Project Astra powers new experiences across Search, the Gemini app, and third-party products. Google collaborates with partners like Samsung to develop Project Astra glasses but hasn't set a launch date yet.

Search Enhancements

This week sees the rollout of AI Mode in Google Search within the U.S., enabling complex queries using an AI interface. It supports sports and finance data queries while offering apparel "try it on" options.

Beam teleconferencing technology uses software and hardware to enable realistic virtual meetings with features like real-time speech translation during Google Meet sessions.

Email Innovations

The introduction of Gemini in Chrome brings a new browsing assistant that aids users in understanding page contexts quickly. Gemma 3n model previews are available now for handling audio, text, images, and videos across devices.

Email innovations include personalised smart replies in Gmail alongside inbox-cleaning features as part of Google's Workspace enhancements.

Wear OS Updates

The Wear OS 6 update introduces a unified font for tiles on Pixel Watches along with dynamic theming that syncs app colours with watch faces. Developers receive design guidelines for better customisation options within apps.

The Play Store receives updates for Android developers with tools for managing subscriptions more effectively alongside topic pages for specific interests like movies or shows (currently U.S.-only).

Subscription Management

A new checkout experience simplifies selling add-ons while subscription management tools gain multi-product checkout capabilities allowing add-ons alongside main subscriptions under one payment method.

The integration of new AI features into Android Studio includes "Journeys," coinciding with the release of Gemini 2.5 Pro model's "Agent Mode" which handles intricate development processes efficiently.