The keyword for the Google I/O 2025 developer was packed on Tuesday. During the session, the CEO of Sundar Pichai and other executives announced a large number of updates and new artificial features (AI). Some of these new capabilities in the Gemini 2.5 series include artificial intelligence models, updates to artificial intelligence mode, expand an artificial intelligence overview, enter the new Google 3D communication platform, and clarify the Android XR platform. In the event that you do not perform the event live, here is a quick tour of everything announced.
Google Ray
The TECH GINT project is now presented as a 3D Etisalat platform. It uses a set of six cameras to capture the user’s video flow from different angles. Then, the artificial intelligence system combines them to convert two -dimensional nutrition into the 3D light field screen. The company also uses head tracking sensors to accurately capture the user at 60 frames (FPS).
Google works with HP to provide the first Google Beam later later this year. Initial devices will only be provided to identify customers. In addition, Google Beam products from the original equipment manufacturers (OEMS) will be provided via InfoComm 2025, which was appointed for the month of June.
Gemini 2.5 promotions
The Gemini 2.5 series also gets many new capabilities. The new deep thinking mode is added to the 2.5 Pro model, which is called augmented thinking mode. The feature is currently under test. The original sound output is also added, which is also to generate expressive and human speech to Gemini 2.5 models via the API.
Google also updates the Flash Gemini 2.5 model with improved capabilities through thinking, weakness, symbol and long context. The model will also be more expensive for use. The developers who use the Gemini applications interface will also get thinking summaries and thinking budget with the latest models.
Artificial intelligence mode in research
Another major point of discussion from the keyword session was artificial intelligence mode in the research. Google is now planning to run AI’s to -tip search with a dedicated gemini 2.5 model. AI’s mode also gets a new deep search mode, a direct search feature that allows the artificial intelligence tool to access the device’s camera, and a new agent to allow users to buy event tickets and reserve a direct date from the interface.
The artificial intelligence mode in the research also gets new features that focus on shopping. Users will now get visually searching for the product they want, and try almost a wide range of clothes by downloading a picture of themselves, and using artificial intelligence agents to track the prices of products and perform purchases automatically. These features will be added later this year.
An artificial intelligence overview expansion
The Mountain view, which is taken by Mountain View, has announced the expansion of an artificial intelligence overview during the keyword. The Acting Search Shot feature will now be available in more than 200 countries and more than 40 languages. With this update, it will support the Arabic, Chinese, Malay and Urdu, which joins the current language options such as the English, Indonesian, Indonesian, Japanese, Portuguese and Spanish.
Android XR works
During the keyword session, Google also showed an experimental offer for the new Android XR platform that works in Jimini. The operating system for the upcoming Samsung project will be Moohan Smart Glasses. The company also works with other partners that can be worn.
This Android XR smart glasses will display a camera, microphone, speaker and display on the glasses. Users will be able to hold hands -free conversations with Gemini, ask them to take pictures and control their smartphone and other connected devices, and more.
Image 4 and see 3 models of Amnesty International
Google has also unveiled the next generation of its image generation model, imagen 4 and VEO 3 video generation model. Imagen 4 now comes with the improvement of text texts and an understanding of the context of text mode, as well as improving image quality and rapid commitment.
With VEO 3, the company adds the ability to generate the original sound, which means that the videos created will display the surrounding sounds, back music as well as dialogues. Both models will be released to the public later this year.
The company also launches a new application for the Acts, which is called Flow. It enhances Imagen, Veo and Gemini to create eight seconds. Multiple clips can also be sewn together to create a longer scene. The application accepts both text and images as claims.
Gemini in Google Chrome
The paid subscribers will be able to reach the Gemini Ai assistant inside Google Chrome. The new Gemini button will allow users to summarize a web page or ask questions about the content. It can also navigate on websites automatically, based on user claims. Artificial Intelligence Assistant can also work on multiple tabs at the same time.
Google has also unveiled a new self -powered tool that can create application interfaces based on text and template claims. Supports App Dobbed Stitch, also supports wire applications, rough graphics, and screen clips for other UI user interface designs. It is currently available as an experience via Google Labs.
Speech translation in Google Meet
The company also adds a new AI feature to Google Meet. The video conference platform will now support the actual time translation and can help speakers with different original languages that speak with a simple delay. Currently, the feature can translate the English and Spanish language. It is currently available to subscribers paid in the experimental version.
Google has superior crying
Finally, Tech Giant also provided Google Ai and Google Ai Ultra plans for a group of Gemini features. The former replaces the Google One Ai Premium plan and will be available for $ 19.99 (1950 rupees in India) per month, while the Google Ai Ultra plan will cost $ 249.99 (about 21,000 rupees) per month. The latter will get all the new features first, provides the limits of the highest rate, and provide 30 terabytes of cloud storage.
