Highlights
- Google introduced new AI subscription plans, Gemini 2.5 Flash and Deep Think, AI-powered tools for Search, Workspace and Project Astra integration.
- AI Mode now provides direct answers, taps into Gmail context, supports shopping and ticket purchases and expands to 200+ countries.
- Wear OS 6 Developer Preview, Jules AI coding assistant, AI-powered filmmaking, Google Play enhancements and XR partnerships.
Caption – Google at I/O 2025. (Image credit – Google)
Google I/O 2025 was all about AI and how it’s going to power everything from Search and Gmail to Chrome, Android and even shopping. While Android didn’t take centre stage during the main keynote, Google still had plenty to share across its products and services.
You can check out the entire keynote below –
Here’s a simple breakdown of all the big announcements.
Gemini
Caption – Google Gemini plans. (Image credit – Google)
Google is rebranding AI Premium to Google AI Pro at the same price of $19.99/month. It gives you more features and higher usage limits than the free tier, with more updates on the way.
Students in Japan, Brazil, Indonesia, and the UK can now get AI Pro free for one school year.
For power users, Google introduced a new AI Ultra plan at $249.99/month, giving you early access to cutting-edge tools.
- Gemini 2.5 Flash is now out of preview, running faster and more efficiently.
- Gemini 2.5 Deep Think, coming soon, will offer stronger reasoning in math, code, and multimodal tasks, but it’s still undergoing safety checks. It’ll launch with AI Ultra first.
- Imagen 4 with improved image and text generation, is now rolling out in the Gemini app.
- Veo 3 adds native audio generation to videos and is also exclusive to AI Ultra in the U.S.
- Gemini Live is getting camera and screen sharing support in the iOS app.
- Soon, Gemini Live will support Calendar, Keep, Maps, and Tasks, helping you with more day-to-day tasks.
Caption – New Gemini announcements made by Google during I/O 2025. (Image credit – Google)
You’ll also be able to upload your own PDFs and images to fuel Deep Research, and what you create in Gemini Canvas, like notes or drafts, can now become a webpage, infographic, quiz, or audio summary.
Gemini can now generate interactive quizzes too.
Google also previewed the latest from Project Astra, showing Gemini controlling Android phones, making calls, and navigating apps.
Gemini and Gemini Live are heading to desktop Chrome (Mac and Windows). At launch, they’ll answer questions about webpages, with full site navigation coming later.
Project Mariner, which turns Gemini into a smart agent that can complete tasks online, will be available to AI Ultra users. This powers the upcoming Agent Mode in the Gemini app.
Search Gets Smarter with AI
AI Mode in Google Search is now rolling out to all U.S. users. It will get new features first before they appear in AI Overviews, and both will run on a custom version of Gemini 2.5.
- AI Mode can soon tap into your Gmail context to deliver personalised results.
- It will also support custom charts and graphs for things like sports scores or finance trends.
- A shopping panel will update live as you refine your search.
- Deep Search, similar to Deep Research, is also coming.
- Search Live will use Project Astra to offer real-time, interactive AI help.
- You’ll be able to say things like “Find 2 affordable tickets for Saturday’s Reds game in the lower level,” and AI will handle the ticket purchase and even book restaurant reservations with your approval.
Caption – AI Mode explained at Google I/O 2025. (Image credit – Google)
The “buy for me” tool is also coming to AI Mode, allowing Google to add items to your cart and checkout via Google Pay.
And yes, you’ll be able to virtually try on clothes by uploading a photo. The AI understands how fabrics fold and stretch on different body types.
AI Overviews are expanding to 200+ countries in 40+ languages.
Google Workspace Updates
Caption – New Google Meet Speech Translation upgrades explained during Google during I/O 2025. (Image credit – Google)
- Gmail is getting Personalised Smart Replies, Inbox Cleanup, and Calendar scheduling tools.
- Google Meet Speech Translation will now feel like having a live interpreter on the call. It’s available now for AI Pro and Ultra users.
- Docs will let you ground AI responses with info directly from your document.
- Imagen 4 is coming to Docs, Slides, Vids, and more.
- Google Vids can now turn a full slide deck into a video with AI avatars that only need a script. Other new tools include Transcript Trim and Balance Sound.
More AI Announcements
Caption – Google Beam at Google I/O 2025. (Image credit – Google)
- Project Starline has been renamed to Google Beam, a new 3D video communication platform. HP is making the first device.
- Google teased more XR glasses demos and announced a bigger partnership with Samsung.
- Google Flow is a new AI filmmaking tool, evolving from VideoFX. It uses Gemini, Veo, and Imagen to let you create scenes with simple text prompts.
- Google introduced the SynthID Detector, a portal where you can upload content to check if it includes Google’s AI watermark.
Tools for Developers
- Jules is Google’s new AI coding agen It “reads your code, understands your intent, and gets to work.” Jules can write tests, fix bugs, and even upgrade dependencies. It’s in public beta with no waitlist now.
- Wear OS 6 Developer Preview is out with a new Material 3 Expressive design.
- Android Studio now has built-in Gemini 2.5 Pro and a new Studio Labs to test AI tools in stable versions.
Google Play Updates
- Devs can now pause buggy app updates before they reach more users.
- The Play Store is getting curated spaces for local and seasonal interests like football in Brazil or cricket in India.
- New topic browse pages will help users discover timely content, like live sports and movies.
- Google Play Collections are expanding with a Travel category and are coming to Brazil, India, Indonesia, Japan, and Mexico.
From deeper AI integration across platforms to a personal AI agent that can shop and organise your day, Google I/O 2025 clearly marks the company’s push into a future where AI touches almost everything you do.
FAQs
Q1. What new AI features did Google announce for Gemini?
Answer. Google introduced Gemini 2.5 Flash for faster AI responses and Gemini 2.5 Deep Think for stronger reasoning in math, coding, and multimodal tasks.
Q2. How is Google making Search smarter with AI?
Answer. AI Mode now provides AI-generated answers, taps into Gmail context, supports shopping and ticket purchases, and expands to 200+ countries in 40+ languages.
Q3. What updates did Google reveal for developers?
Answer. Google announced Wear OS 6 Developer Preview, Jules AI coding assistant, AI-powered filmmaking tools, and new Google Play updates for app discovery and curation.
Also Read: