On Tuesday, Google unveiled Gemini 2.5 Pro Preview (I/O edition), an enhanced iteration of its leading Gemini 2.5 Pro AI model, asserting its superior performance across several prominent benchmarks.

The Gemini 2.5 Pro Preview (I/O edition) is accessible through the Gemini API, alongside Google’s Vertex AI and AI Studio platforms, and is priced equivalently to the Gemini 2.5 Pro model it largely supersedes. Furthermore, it is integrated into Google’s Gemini chatbot application for both web and mobile interfaces.

The model's launch precedes Google's yearly I/O developer conference, thus its "I/O edition" label, where Google is predicted to reveal numerous models, alongside AI-driven tools and platforms. The corporation is intensely vying for recognition and market dominance in the fiercely competitive AI arena; competitors such as OpenAI and xAI are on the verge of launching models anticipated to demonstrate exceptional performance.

Google says that Gemini 2.5 Pro Preview (I/O edition) is much better at coding and making interactive web apps. The company also states that the model is improved for tasks like changing code to reach a certain goal (code transformation) and editing code.

In a blog entry, Google indicates that Gemini 2.5 Pro Preview (I/O edition) tops the WebDev Arena Leaderboard, a standard for evaluating a model's aptitude for crafting visually appealing and practical web applications. Gemini 2.5 Pro Preview (I/O edition) also exhibits cutting-edge capability in video analysis, reaching a score of 84.8% on a widely used benchmark, VideoMME.

In a blog post, Google said that the new version of Gemini 2.5 Pro will be better for developers who already use it. It will improve how well they can code and also fix important issues developers mentioned, like having fewer errors when calling functions and making function calls happen more often. Google also said the model naturally prefers nice-looking web design but can still be controlled.