Google is betting everything on AI and making a radical change to remain the favorite search engine.

The new features were revealed during the official presentation of just over two hours at Google's annual conference. This year, the focus was on showcasing how AI is driving Google's mission to make information universally accessible and useful for everyone, both users and developers.

“There’s more intelligence available, for everyone, everywhere. And the world is responding, adopting AI faster than ever. All this progress means we’re in a new phase of AI platform change, where decades of research are now becoming reality for people, businesses, and communities around the world,” said Sundar Pichai , CEO of Alphabet and Google.

This common thread served to showcase AI functions in online shopping, email writing, assistance when using virtual reality headsets, and even simultaneous translation when communicating via Google Meet with people who don't speak our language.

Many of the presentations were done with live demonstrations, which included anxious expressions on the part of the presenters and some minor glitches that ended up helping to clear up doubts about whether it was a staged project or actually being tested live.

Google is improving its most advanced models—the Gemini 2.5 Flash and Pro—which will soon be generally available. Gemini 2.5 Pro now includes Deep Think, a new reasoning mode designed to improve performance on complex tasks like advanced math and programming.

Gemini is being extended to become a global model capable of planning and imagining new experiences by simulating aspects of the world. Project Astra's live capabilities, such as video compression, screen sharing, and memory, are also being integrated into Google products like Gemini Live and Search Live.

AI Mode in Finder is now available to all users in the US, providing a more interactive experience powered by Gemini's most advanced models. At the same time, AI Overview is now available in over 200 countries and territories, and in over 40 languages.

Google AI Ultra is here, a new premium subscription plan that offers access to its most capable AI models, higher usage limits, and exclusive features—starting at $249.99/month in the US, with global expansion coming soon.

Gemini in Workspace continues to evolve, adding features like personalized smart replies in Gmail, real-time voice translation in Meet, and expanded support for Docs, Videos, and Slides—helping people write, communicate, and collaborate more effectively.

Google unveils Image 4 and Veo 3, its latest generative models for creating high-quality images and videos, along with Lyria 2 for music generation.

Google Beam, the evolution of Project Starline, is the new AI-powered platform for 3D communication. Also launching is a preview of Speech Translation, enabling high-fidelity, real-time voice translation between languages, preserving tone and nuance for more natural global conversations. Creators can now also explore new storytelling formats with Flow, a new AI-powered filmmaking tool designed to bring ideas visually to life.

The new AI Mode shopping experience leverages Gemini AI and Google's Shopping Graph to offer personalized navigation, virtual try-ons for billions of items, and an assisted shopping feature that can purchase items on your behalf through Google Pay when the price is right, always under your control.

It's a standalone, asynchronous coding assistant that integrates directly with your existing repositories and handles tasks on your behalf, and is not available in public beta.

Learn how Android XR and Gemini will be coming to devices like glasses and headsets, and which partners will be announced.

The highlight of the presentation is the large number of new features Google announced, especially the integration of Artificial Intelligence into the ecosystem of all its services. This will allow new AI services to be used in search engines, email, and the web for shopping. It also gives a boost to virtual reality headsets and voice assistants.

However, not everything is wonderful. Many of the announcements don't have a precise release date, and others will only be available to users in the United States.

These features, especially those related to online shopping, have always worked to a limited extent in Argentina and Latin America. It remains to be seen whether the universality of AI developments will help improve the company's services in more countries.