Google I/O 2025: Top 15 Major Reveals

Google has just concluded its major keynote at I/O 2025. As anticipated, the event was packed with AI-related announcements, from enhancements to Google's image and video generation models to fresh capabilities in Search and Gmail.

However, there were a few unexpected reveals, including a new AI filmmaking application and an upgrade to Project Starline. If you missed the live presentation, you can catch up on all the developments in the summary below.

Google's AI Search Mode Expands to All Users

Google has revealed that it is launching AI Mode, a new tab enabling web searches through its Gemini AI chatbot, to all U.S. users starting this week.

This summer, Google will trial new functionalities within AI Mode, including deep search and tools for generating charts related to finance and sports queries. Additionally, shopping within AI Mode will become available in the coming months.

Project Starline Rebranded as Google Beam

Project Starline, which originated as a 3D video chat booth, is evolving significantly. Now rebranded as Google Beam, it will soon debut in an HP-branded device featuring a light field display and six cameras to produce a 3D image of the person you are video calling.

Companies such as Deloitte, Duolingo, and Salesforce have already announced plans to equip their offices with HP's Google Beam devices.

Imagen and Veo Receive Major Enhancements

Google has introduced Imagen 4, the newest iteration of its AI text-to-image generator, which offers improved text rendering and support for exporting images in various formats, including square and landscape. The next-generation AI video generator, Veo 3, allows for synchronized video and audio generation, while Veo 2 now includes features like camera controls and object removal.

Google Unveils AI Filmmaking Application

Alongside updates to its AI models, Google is launching a new AI filmmaking app named Flow. This tool leverages Veo, Imagen, and Gemini to produce eight-second AI-generated video clips from text prompts and/or images. It also includes scene-building tools to combine clips into longer AI-generated videos.

Gemini 2.5 Pro Introduces Enhanced Reasoning Mode

The experimental Deep Think mode is designed for complex queries involving math and coding. It can evaluate multiple hypotheses before responding and will initially be available only to a select group of trusted testers.

Google has also made its Gemini 2.5 Flash model accessible to all users through the Gemini app and is implementing improvements to the cost-effective model in Google AI Studio ahead of a broader release.

Xreal Unveils Project Aura Prototype

Xreal and Google are collaborating on Project Aura, a new set of smart glasses that utilize the Android XR platform for mixed-reality experiences. While details are still limited, the glasses will feature Gemini integration, a wide field of view, and apparent built-in cameras and microphones.

Google is also partnering with Samsung, Gentle Monster, and Warby Parker to develop additional Android XR smart glasses.

Google's Experimental AI Assistant Becomes More Proactive

Project Astra could already utilize your phone's camera to identify surrounding objects, but the latest prototype enables it to perform tasks autonomously, even without direct instructions. The model can initiate conversation based on its visual input, such as highlighting an error in your homework.

Gemini Integration Coming to Chrome

Google is embedding its AI assistant directly into Chrome. Starting May 21st, Google AI Pro and Ultra subscribers will be able to use the Gemini button to clarify or summarize information from webpages and navigate sites automatically. Google plans to enable Gemini to operate across multiple tabs simultaneously later this year.

Google's New AI Ultra Plan Priced at $250 Monthly

Google is introducing a new "AI Ultra" subscription that provides access to its most advanced AI models and increased usage limits across applications like Gemini, NotebookLM, Flow, and others. The subscription also includes early access to Gemini in Chrome and Project Mariner, which can now handle up to 10 tasks concurrently.

Search Live Enables Real-Time Camera Discussions

In relation to Project Astra, Google is launching Search Live, a feature that integrates capabilities from the AI assistant. By tapping the new "Live" icon in AI Mode or Lens, you can engage in two-way conversation with Search while displaying your camera feed.

Following last month's decision to make Gemini Live's screensharing feature free for all Android users, Google has announced that iOS users will also receive free access.

Google's New AI Tool Generates App Interfaces

Google has introduced Stitch, a new AI-powered tool capable of generating interfaces based on selected themes and descriptions. You can also input wireframes, rough sketches, and screenshots of other UI designs to influence Stitch's output. This experiment is currently available on Google Labs.

Google Meet Adds AI Speech Translation

Google Meet is introducing a new feature that translates your speech into your conversation partner's preferred language almost instantly. Initially supporting only English and Spanish, the feature is rolling out in beta to Google AI Pro and Ultra subscribers.

Gmail's Smart Replies to Incorporate Inbox Information

Gmail's smart reply feature, which uses AI to propose email responses, will now draw information from your inbox and Google Drive to craft replies that better reflect your writing style. The feature will also consider your recipient's tone, suggesting more formal responses in conversations with your manager, for instance.

The enhanced smart replies will be available in English on web, iOS, and Android when it launches via Google Labs in July.

Google Embraces AI-Powered Shopping

Google is testing a new functionality that allows you to upload a full-body photo to visualize how shirts, pants, dresses, or skirts might fit. It employs an AI model that comprehends human anatomy and clothing details.

Google will also soon enable shopping within AI Mode, along with an "agentic checkout" feature that can complete purchases on your behalf.

Google Chrome to Assist with Updating Compromised Passwords

If Chrome identifies that your password has been compromised, Google states the browser will soon be able to create a strong replacement and update it automatically on supported websites. This feature launches later this year, and Google confirms it will always request consent before modifying your passwords.