Google updates Gemini 2.5 Pro model for coders

Google has updated its Gemini 2.5 Pro AI model with stronger coding capabilities, hoping developers would begin building with the model before the Google I/O developer conference later this month.

Released May 6, Gemini 2.5 Pro Preview (I/O edition) is accessible in Google AI Studio tool and in Vertex AI.

Commenting on the update, Google said developers could expect meaningful improvements for front-end and UI development alongside improvements in fundamental coding tasks including transforming and editing code and building sophisticated agentic workflows. Developers already using Gemini 2.5 Pro will find not only improved coding performance but reduced errors in function calling and improved function calling trigger rates, the company said.

The Gemini 2.5 Pro model shines for use cases such as video to code, due to the model’s “state-of-the-art video understanding,” and offers easier web feature development, due to the model’s “best-in-class front-end web development,” Google said.