GOOGLE I/O 2025: See all the new artificial intelligence announcements
Table of Contents
Expected Google I/O 2025 finally happened and several new features that should change our routine were announced. At its traditional event, the company revealed new features for video creation, changes to the search tool Google Chrome and the global launch of Gemini Live with camera, which was being tested since the beginning of the year. Check out everything that was announced right now.
What's New in Google Search
O Google Chrome has received several new features to boost its search function with the support of artificial intelligence. The company has taken advantage of its new Gemini 2.5 language model to allow the browser to not only do all the search work, but also to help with online shopping.
AI Mode
Tested since March this year, the AI Mode has finally been released to the general public in the United States. The idea is that Gemini will do all the research work and bring you comprehensive results based on what you like and have previously searched for. AI Mode will be in a new tab on the Google Search screen so that, if you prefer to use the traditional search, the user still has this power of choice.
A Google highlighted that his new AI Mode It may also use the data used in Google applications, such as suggesting restaurants with the food you like, favorite places and even showing what you can do based on what was discussed in other applications. The company emphasizes that “Personalization”, as the function was called, can be disabled if the user wishes.
O AI Mode works in a similar way to artificial intelligences like Manus AI e ChatGPT They already do, but it should be used more because it is integrated into the browser widely used in people's daily lives.
Still on AI in Search, the company highlighted that AI Overviews, that AI-generated summary that appears at the top of general search queries, will now be available in over 200 countries and with support for over 40 languages.
Try On (Clothes fitting room)
With the function name Try On, this is the feature that will allow you to try on clothes virtually, and you only need to send a full photo of your body to get a preview of the clothes you are thinking of buying. The new feature began testing today in the US and should be released to other countries later.
Google says the feature uses an AI model that “understands the human body and the nuances of clothing — how different materials bend, stretch, and drape on different bodies.”
A purchasing agent, powered by Gemini, was also announced by the company during the Google I/O 2025. Basically, in addition to tracking prices, the user can send data such as the item's color, size and the amount they wish to spend. When the indicated amount is found on the internet, the purchasing agent will finalize the purchase, previously authorized, via Google Pay.
Gemini Live for everyone
Released for testing to users of Galaxy S25, Gemini Live, which uses the camera to understand the context of the scene and can even be used as support in research, was also released to all people who have the app. Gemini on your smartphone with Android ou iOS.
The idea here is to guarantee an evolution to those who already use it. Google Lens in search, but with a more intelligent language model that can respond in real time, as if you were talking to your personal assistant.
Soon, the Gemini Live will gain integration with applications Google for appointments (in the Agenda), indicate paths in the Maps and also put notes on the Google Keep. This item will be available as a shortcut in the corner of the screen, as you can see below:
Speech Translation: Real-time translation in Google Meet
Through the Google Deep Mind, it will now be possible to access real-time translation in Google Meet. The feature is designed to perform audio conversion and, according to the company, it is performed with low latency. It will basically still allow the voice of the person being translated to be heard, but with a slight overlap, generated by artificial intelligence.
The resource Speech Translation will be tested starting today, May 20, in the US, for subscribers of Gemini plans. The company mentions that, initially, it works with Spanish and English, but support for languages such as Italian, German and Portuguese will be available in the “coming weeks”.
Updates on Deep Research
Google's famous AI researcher, the Deep Research (which, by the way, has already been subject matter here at showmetech), now supports sending documents from the user's computer. With this, in addition to being able to read, analyze and create summaries of what was sent, the AI can use the sent document as a basis for new research. The new feature will be available to all users by the end of this week.
Soon, the Google Deep Research There will also be a button that will allow access to Google Drive, so that files saved in your Google account can be accessed directly from there, similar to what happens with Google Docs. A release date for this has not yet been announced, but the company says it should be released between July and September.
Image 4: New version of AI that creates images
Known for being one of the best options for generating images with texts, Image, Google’s AI focused on creating photos, has received a new version. The company says that this new version can now generate “even more realistic images, with more vivid colors and without presenting those errors already known to AI users”.
Another new feature that comes with the Imagen 4 is the stylization of fonts, also without typos. The company cited an example of an invitation to a dinosaur exhibition that can have fossil-style fonts in different models. It also allows you to create realistic and abstract images in up to 2K resolution, in addition to supporting multiple content formats.
Imagen 4 is our best text-to-image model yet, with photorealistic images, real-time speed, and sharper clarity — to bring your imagination to life.
Google on new image model released today.
O Imagen 4 is available on the Gemini app, Whisk, Vertex AI and Google Workspace applications (Docs, Slides, Vids).
VEO 3 + Flow: creating videos with audio
Direct competitor of Sora from OpenAI, I SEE 3 stands out for having support for creating videos based on text commands, with 2K resolution and up to eight seconds in length. Its biggest feature is the possibility of creating videos with sound, such as this video of a stand up, including the joke and laughter from the audience. Here are some examples:
During its event, the company highlighted that the I SEE 3 can be used in many different situations: animations, videos with people and more. One of the most surprising examples in Google I/O 2025 It was the animation of an owl, during the night, in which we heard the noises of its wings.
already about the Flow, this is another text-based video creation platform, more focused on creating content up to 8 seconds long. The idea here is that users can upload static images and see them come to life, including through new contexts.
Both I SEE 3 as for Flow will be available initially in the United States through plans announced today at the event. The plan Gemini AI Pro, which already exists and currently costs R$96,99, now supports the platform Flow. Anyone who wants to use the platform I SEE 3, you need to pay US$249,99 for the plan Gemini AI Ultra (R$1.415 in direct conversion per month), which includes the creation of static videos with sound.
Android XR and Samsung smart glasses
Closing the news, the Google I/O 2025 was also marked by news from Android XR, operating system for VR headsets and smart glasses, announced in late 2024.
In addition to demonstrations of the operating system that is expected to debut in a Samsung VR glasses (whose initial name is Project Moohan) and will be available by the end of the year, it was also announced that a new smart glasses, along the lines of the Ray Ban Meta, produced by Samsung. More details about the upcoming device have not yet been revealed, but we will update you as more news is released.
A partnership with companies Gentle Monster e Warby parker was also announced. The idea here is that both brands will take care of the style, such as colors, frames and other visual details of the future Google and Samsung smart glasses.
What do you think of all the news from Google I/O 2025? Which one are you most excited to try out? tell us Comment!
See also other features
With some info: Google Blog l The Verge (1 e 2).
Text review done by Jaime Ninice on 22/05/2025.
Discover more about Showmetech
Sign up to receive our latest news via email.
