Everything Google announced at I/O 2025: Gemini, Android, more


At I/O 2025, Google announced an absolute torrent of new features, powered by Gemini and AI, across its biggest products and services that will soon be available to users. We’ve compiled all the consumer-facing announcements and notable developer developments made at the event below.

Updating…

Links will go live over the course of the keynote

Gemini

AI Premium is being renamed to Google AI Pro at the same $19.99 per month. You get higher limits than the free tier, and Google will continue to add new features here.

Advertisement – scroll for more content

  • Free AI Pro for a university school year is going live in Japan, Brazil, Indonesia, and the United Kingdom.

Google announced a new Google AI Ultra tier that costs $249.99/month to get early and maximum access to the bleeding edge. 

Gemini 2.5 Flash is out of preview with performance improvements and efficiency gains. Google touts how 05-20 is only second to 2.5 Pro on the LM Arena Leaderboard. 

Gemini 2.5 Deep Think will be a new “enhanced reasoning” mode that leads across benchmarks for Mathematics, Code, and Multimodality. It’s undergoing additional safety evaluations before launching, and will be available to Google AI Ultra subscribers. 

Imagen 4 with better detail and text outputs is rolling out the Gemini app. 

Veo 3 features native audio generation to accompany video. It’s available with Google AI Ultra in the US.

Gemini Live camera and screen sharing is coming to the iOS app. 

In the “coming weeks,” apps/Extensions support is coming to Gemini Live starting with Calendar, Keep, Maps, and Tasks. 

You can now upload your own PDFs and images to inform Deep Research. It gets combined with public data.

What you write in Gemini Canvas can now be turned into a Web page, Infographic, Quiz, Audio Overview, and more. 

Gemini can now make interactive quizzes.

Google showed off the latest Project Astra capabilities, like controlling an Android phone, navigating apps, and making calls.

Gemini and Gemini Live are coming to desktop Chrome (Mac + Windows). At launch the focus is letting you ask questions about the current page, with website navigation coming later this year. 

Project Mariner will be available for Google AI Ultra subscribers.

Gemini app users will get an “Agent Mode” powered by Mariner that can accomplish tasks by browsing the web.

Google Search 

At I/O 2025, Google announced that AI Mode is now rolling out to all US users. New features will debut first there before trickling down to AI Overviews. Both will soon use a custom version of Gemini 2.5. 

AI Mode can soon use your personal context from Gmail for “tailored results” that reflect your preferences. It will also get the ability to create custom charts and graphs to visualize sports and finance queries.

Shopping capabilities are coming to AI Mode with a panel that updates with relevant products as your refine your search  

Deep Search is coming to AI Mode. It sounds similar to Gemini’s Deep Research.

Search Live will be using Project Astra capabilities to offer a Gemini Live-esque experience 

Project Mariner’s agentic capabilities are coming to AI Mode to let you ask questions like: “Find 2 affordable tickets for this Saturday’s Reds game in the lower level.” Google can purchase the event ticket and make the restaurant reservation or local appointment with your approval. 

Another use of this agentic capability is “buy for me” when shopping to add items to the cart and checkout with Google Pay. 

Google will let you try on clothes virtually when you upload a picture of yourself. It takes into account how materials fold and stretch on different bodies 

With the latest expansion, AI Overviews are coming to 200+ countries in over 40 languages. 

Google Workspace

Gmail is getting Personalized Smart Replies, Inbox Cleanup, and integrated Calendar appointment scheduling.

Google Meet Speech Translation is like having a human interpreter on the call. It’s available to test with AI Pro and Ultra starting today. 

Google Docs will let you ground Gemini side panel responses.

Imagen 4 is coming to Google Docs, Slides, Vids, and other Workspace apps.

Google Vids will let you turn a whole slide deck into a video, AI Avatars just require you to upload a script. Other features coming are Transcript Trim ad Balance Sound 

More AI at Google I/O 2025

Project Starline is now Google Beam, an “AI-first 3D video communication platform.” HP is making the first device.  

Besides sharing more demos for Android XR glasses, Google announced an expanded Samsung partnership

Google Flow is an “AI filmmaking tool” built on Veo, Imagen, and Gemini models. Described as an evolution of VideoFX, the latter makes possible natural-language prompting, with camera controls and a scene builder.  

Google announced the SynthID Detector “verification portal to help people identify Al-generated content.” You upload content and Google returns whether the “entire file or just a part of it has SynthID [watermark] in it.”

Notable developer announcements

Jules is described by Google as a “true coding agent” that “reads your code, understands your intent, and gets to work” (in the background). Specifically, Jules can write tests, build new features, fix bugs, and bump dependency versions. It’s available in public beta with no waitlist worldwide.  

Wear OS 6 Developer Preview: Available as an emulator with updated Material 3 Expressive design guidance 

Android Studio: The built-in Gemini integration is now powered by 2.5 Pro. Google is making it easier to test AI features in the stable channel with Studio Labs 

Google Play

  • App developers can now “halt fully-live releases” that have an issue. The update won’t be offered to users that haven’t installed it yet 
  • Google’s push for a “content-rich” Play Store is going well with curated spaces that “celebrate seasonal interests like football (soccer) in Brazil and cricket in India, and evergreen interests like comics in Japan.” Expect more categories and locations this year 
  • New topic browse pages can “present timely, relevant, and visually engaging content on specific topics, like live sports, movies, and shows. 
  • Google Play Collections are adding a new Travel category and coming to Brazil, India, Indonesia, Japan, and Mexico. Content that appears there will soon also appear in the Play Store.

FTC: We use income earning auto affiliate links. More.



Source link

Previous articleAndroid TVs Now Have Those Annoying App Review Popups