Google's Gemini 2.5 Pro update makes the AI model even better at coding
Google has released early access to a version of Gemini 2.5 Pro with significantly improved coding capabilities. The model is reportedly better at coding tasks such as building interactive web apps, code transformation, code editing, and creating agentic workflows. It was scheduled to debut at Google I/O, but Google released it early in response to positive feedback. The updated Gemini 2.5 Pro is placed at the top of the WebDev Arena Leaderboard and scored 84.8% on the VideoMME benchmark.