Google Unveils Enhanced AI Model Gemini 2.5 Pro with Advanced Web App Development Features

Google has launched an updated version of its flagship AI model, Gemini 2.5 Pro. The new version features «significantly enhanced» programming capabilities, specifically tailored for «creating convincing interactive web applications.»

«We have just released Gemini 2.5 Pro (I/O edition), our most advanced model yet, which excels at coding,» the company announced.

«Now, you can develop interactive web apps in Canvas with fewer prompts needed.»

«Visit https://t.co/BCZTezi8r9 and select ‘Canvas’ in the prompt bar to give it a try, and let us know what you think… pic.twitter.com/gjDpf5nkRE»

The Gemini 2.5 Pro Preview (I/O edition) excels in code transformation—modifying code snippets to achieve desired outcomes. It is accessible via the Gemini API, Google Vertex AI platform, and AI Studio, with pricing unchanged from Gemini 2.5 Pro.

The release coincides with the annual Google I/O developer conference (hence the name «I/O edition»), where the company will present a series of new solutions.

Gemini 2.5 Pro Preview (I/O edition) currently leads the WebDev Arena Leaderboard, a benchmark that evaluates a model’s ability to produce aesthetically pleasing and functional web applications.

Additionally, it showcases top-tier performance in video comprehension, achieving a score of 84.8% in VideoMME.

The Gemini 2.5 Pro is particularly proficient in frontend development.

«For developers already using Gemini 2.5 Pro, this new version not only enhances coding performance but also addresses key feedback from developers, such as reducing errors during function calls and improving function call accuracy,» Google mentioned in a blog post.

It’s worth noting that in May, Google introduced the capability to upload and edit images within the Gemini chatbot application.

Previously, the company integrated the Veo 2 video generation AI model into Gemini for users subscribed to Gemini Advanced.