Google launches Gemini 2.5 Pro Preview with upgraded coding capabilities ahead of I/O

Google has officially launched Gemini 2.5 Pro in preview, following a month of experimental rollout. This update delivers key advancements for developers, expanding coding capabilities with stronger tools for front-end and UI development. Developers will notice improved support for core coding tasks, including code transformation, editing, and the creation of sophisticated agentic workflows. Gemini 2.5 Pro now sits at the top of the WebDev Arena leaderboard, indicating human preference for its ability to produce aesthetically pleasing and functional web applications.

Technical benchmarks show further strengths: the model achieves state-of-the-art video understanding, earning an 84.8% score on the VideoMME benchmark. This combination of video comprehension and enhanced code generation enables user experiences not possible with earlier models. For users wanting to integrate Gemini 2.5 Pro, the model can be accessed via the Gemini API in Google AI Studio, or through Vertex AI for enterprise use. Additionally, the release addresses key developer feedback, reducing errors in function calling and improving how function calls are triggered, strengthening reliability for ongoing projects.