Google has unveiled a preview version of its updated Gemini 2.5 Pro AI, named the I/O edition, ahead of its annual developers event. This enhanced model can now be accessed through the Gemini API as well as Google’s Vertex AI and AI Studio with no price change from its predecessor.
The Gemini 2.5 Pro Preview is now also offered inside the Gemini app for both web and mobile, granting more users direct access to its advanced features. Google is intensifying efforts to outpace competitors like OpenAI and xAI, both of which are developing impressive new AI models set for release soon.
Expanded Coding and App Development Abilities
With this latest release, Google highlights significant improvements in building interactive web applications and coding tasks. Capabilities in code transformation and editing have been sharpened, helping users modify code more efficiently to achieve specific development goals.
The model outperforms others in the WebDev Arena Leaderboard, reflecting its talent for crafting attractive, functional web applications. Gemini 2.5 Pro Preview also excels in video comprehension, achieving an impressive 84.8 percent on the VideoMME benchmark.
In an official blog update, Google explained that developers making the switch from the previous Gemini model will benefit from smoother coding experiences. Notable upgrades include reductions in function calling errors and more responsive triggering of function calls.
By default, the new model shows a preference for aesthetically pleasing web design without sacrificing user steering or customization. As Google’s developer conference approaches, the I/O edition sets the stage for forthcoming advancements in AI-powered platforms and tools.
