Google says its updated Gemini 2.5 Pro AI model is better at coding

On Thursday, Google unveiled an enhanced version of its Gemini 2.5 Pro AI model, touting improvements in its coding abilities. This update, described as an 'updated preview,' builds on the previous enhancements made to Gemini 2.5 Pro just a month ago. The tech giant announced that the model will be fully available in a few weeks, while developers can access it now via their AI platforms, including AI Studio, Vertex AI, and the Gemini app. In a blog post, Google stated, 'Gemini 2.5 Pro continues to excel at coding, leading on challenging coding benchmarks.' The model has demonstrated outstanding performance in rigorous tests that assess its mathematical, scientific, and reasoning skills. Additionally, Google highlighted that it has incorporated user feedback from the earlier 2.5 Pro release, resulting in a model that is 'more creative with better-formatted responses.'