Google continues to develop and improve its Gemini AI assistant. After the release of the model Gemini 2.0 Flash optimized for daily activities and other tasks, the company announced about the launch of Gemini 2.5 Pro Preview for developers. Interestingly, it was announced ahead of schedule.
Initially, the new Gemini 2.5 Pro was planned to be presented during the Google I/O conference on May 20-21. However, due to the extremely high interest of the developer community, the company decided to unveil the model earlier.
«Today we’re providing early access to Gemini 2.5 Pro Preview (I/O edition), an updated version of 2.5 Pro that significantly improves coding capabilities, especially for creating immersive, interactive web apps,» Google said.
The new version of Gemini 2.5 Pro Preview was developed based on the feedback from the first release Gemini 2.5 Pro. In the WebDev Arena ranking (a human-based test that measures how well AI creates attractive and user-friendly web applications), the updated model outperformed the previous one by 147 points on the Elo scale.
Google notes that Gemini 2.5 Pro understands code well and is able to reason logically, so it is suitable for developers. In addition, the model has strong video understanding capabilities: it scored 84.8% on the VideoMME test.
Gemini 2.5 Pro is already running in innovative tools such as the Cursor code agent, and is also used in partnerships with companies such as Cognition and Replit that explore new opportunities for developers. Silas Alberti from Cognition said:
«It was like working with a more experienced programmer — the model made the right assumptions and chose good abstractions».
Replit’s Michele Catasta added that they consider Gemini 2.5 Pro «to be the best advanced model in terms of the feature-to-latency ratio». In its blog post, Google also provided examples of practical use of the model. For example, it can turn YouTube videos into full-fledged educational applications, generate stylish UI components, and create web applications from simple ideas.
Developers can already use the new version of Gemini 2.5 Pro via Google AI Studio or Vertex AI. If you have used the previous version, you will be automatically redirected to the latest version — no additional steps or payments are required.
Previously, Google she added to the Gemini app with its experimental artificial intelligence model that can «reason» and explain how it answers complex questions.
Source: neowin
Spelling error report
The following text will be sent to our editors: