Google I/2025: Great change in Gemini 2.5! AI got the human estimate and the strength of Deep Think
Samira Vishwas May 22, 2025 12:25 PM

The Google I/2025 Event was held for two days as May 2 and today, May 21. The company has introduced several tools so far in this event. The company had announced that the experience of users will change completely after this event. Because new tools are designed to give users a fun and better experience. Some of the tools presented in the event will be made available to users next month. Some tools will be made available to users later this year.

Google I/2025: Beam’s announcement in the event, the dream of 3D video conferencing will soon come true

All New Deep Think Advance Reasoning Mode Regards

The company is now adding another new update to the updates and tools introduced by the company. The company has also introduced new updates for Gemini 2.5 AI in this event. In the Google I/O 2025 event, several strong features have been announced for the Artificial Intelligence Model Gemini 2.5. Tech giants have introduced an adjacent renewing mode called All New Deep Think for Chatboat, which will soon be launched for Gemini 2.5 Pro model. This feature can be considered as Chattvpt’s ‘Think for Longer’ features. The company is ready to launch this new feature to compete with their rival Chatgpt, but it would not be possible. (Photo courtesy – Pinterest)

AI estimated estimates

Not only that, the company has also added a Native audio output feature to the Gemini model. This is a new and adapt technology, which gives AI more natural and the strength to speak like humans. This feature will be made available through Live API. In particular, users will be able to control tone, pronunciation and style in their way in this new feature.

New Tools presented for developers too

In addition, the company has also brought new capabilities for Google developers. In this, the company has added a feature called Thot Summer. Simply put, the AI ​​model will now show a detailed summary of its thinking process with an answer. Not only that, Google has recently introduced an updated version of Gemini 2.5 Pro, which has been given better coding capabilities. This model was at the top in the benchmark leaderboard such as Webdev Arena and LMarena.

Google I/2025: Advanced AI Image and Video Generation Tools performed in Events, now Automatic generating Audio

Deep Think answered better

New Deep Think Mode Gemini 2.5 Pro gives Pro’s ability to think in many cases before answering any question. The company says that the use of new research technology will be more beneficial than older models. During the test, it was also revealed that Gemini 2.5 Pro scored 49.4 percent on a difficult math benchmark like 2025 Uamo.

© Copyright @2025 LIDEA. All Rights Reserved.