Google buffs Gemini Pro just before its I/O show
Google has launched Gemini 2.5 Pro Preview (I/O edition), an updated version of its flagship Gemini 2.5 Pro AI model, ahead of its annual I/O developer conference. The new model is available via the Gemini API, Google’s Vertex AI and AI Studio platforms, as well as in the company’s Gemini chatbot app for web and mobile devices.
The Gemini 2.5 Pro Preview (I/O edition) boasts “significantly” improved capabilities for coding and building interactive web apps. It leads the WebDev Arena Leaderboard, a benchmark that measures a model’s ability to create functional and aesthetically pleasing web apps. The model also achieves state-of-the-art performance in video understanding, scoring 84.8% on the VideoMME benchmark.
Video: Demis Hassabis, Co-Founder & CEO of Google DeepMind
Google claims that Gemini 2.5 Pro Preview (I/O edition) is particularly adept at tasks such as code transformation and code editing. For developers already using Gemini 2.5 Pro, the new version promises to improve coding performance, reduce errors in function calling, and improve function calling trigger rates.
Image: Google
The launch comes as Google competes fiercely with rivals like OpenAI and xAI in the AI race. The company’s I/O conference is expected to feature a host of new models, AI-powered tools, and platforms. Gemini 2.5 Pro Preview (I/O edition) is priced the same as the Gemini 2.5 Pro model it replaces.
Google’s blog post highlights the model’s capabilities, stating that it has “a real taste for aesthetic web development while maintaining its steerability.” The company is positioning Gemini 2.5 Pro Preview (I/O edition) as a top contender in the AI landscape, with its improved coding capabilities and performance.