In 2023, Microsoft Chief Executive Officer Satya Nadella famous was famous that he wanted to dance Google. The reference is that – he believed – Redmund found a way to move forward in artificial intelligence (especially general AI) and therefore, the discovery was ready to take Google’s long -standing dominance in the market. He hoped that better competition would push Google to innovate more continuously and come out with better products.

On May 20, 2025, Google earned money from some of its modern success inventions – not one, not two, not two, but a pack slate of AI product announcements, including new business models. In addition, breaking from the convention, Google is planning to release this capacity widely, so that it can be accessible to many more people than before. This was a separate I/O, where Google spoke less to Android and Google TV, and more about how Mithun would run experiences on equipment in the coming days and years.

Google was dropped at I/O 2025, it is not easy to filter the length and scope of each announcement best, but we will let it go anyway: to briefly tell the barrage of information that you really steal the show.

Without further movement, here are five of the largest headline-gabbers from Google I/O 2025:

Gemini 2.5 with deep think mode: Gemini 2.5 Pro, which is the latest and most powerful version of Google’s large language model, is getting advanced logic abilities as a new mode, called deep think. Essentially, this will allow Google to dedicate more computational resources to LLM and answer more complex and wide ways such as more complex and coding challenges. This will make AI so smart that it will be able to detect several hypotheses before reaching a solution. In other words, it is receiving a brain to think and argue like humans.

AI mode in google search: You can think of AI mode as the development of the next stage or Google search, where using Gemini, Google will allow you to ask more complex questions for a long time and achieve direct insights within the AI-operated summary and search results. Google says that the idea is how we want online information at the most fundamental level, going away from the interactions going away from the keyword-based query, which are more natural, like how you will ask another human being.

Android XR and Android XR Glasses: Google is taking another swing on smart glasses. After the failure of Google Glass, the company is once again immersing its toes into a platform with a brand-nine operating system called Android XR, which will run both wearing equipment like Samsung’s project muhhan and smart glass, which will be partner Will be made in The idea of ​​putting Mithun on your face and let it give you useful information such as in front of you or at the most basic level, read your email.

4 imagene with veo 3 and flow: Google has released significant upgrades in its general media model. While the imagene promises better texture and text generation for 4 images, with VO3, you can generate videos and audio together. “Flow” is a new AI film production tool that is released simultaneously that combines these abilities, provides the ability to expand characteristics and scenes such as character and visual consistency, which gives the creators both power and flexibility to create great materials with AI. Some people say, this is the end of Hollywood as we know. In contrast, it is being seen as a big boon for game studios.

Project Estra Integration with Gemini Live: For Gemini, Google’s vision is beyond being a chatbot. It wants it to become a “world model” who can plan, imagine new experiences and follow aspects of the world. To do this, it is integrating live capabilities in the Gemini app from its research prototype project Estra to provide real -time visual aid using the phone’s camera. The system has also been extended to detect emotions in the user’s voice and to react wisely, ignoring background interactions.