Google launches Gemini 2.5 pro preview 

Just ahead of its annual developer conference, Google I/O 2025, tech giant Google has unveiled a powerful new version of its flagship AI model: Gemini 2.5 Pro Preview. Dubbed the “I/O edition,” this early release is designed to give developers a head start in exploring its enhanced capabilities—particularly in the realm of coding.

Demis Hassabis, CEO of Google DeepMind, announced the launch on social media platform X, describing Gemini 2.5 Pro Preview as “the best coding model we’ve ever built.” He highlighted its strength in developing interactive web applications and shared a demo video showcasing how the model can be used to prototype new ideas quickly and effectively.

According to Google, the new model offers significant advancements across multiple coding domains. It excels in front-end and UI development, code transformation and editing, and the design of agentic workflows. These upgrades aim to reduce development time and improve the overall efficiency of building innovative software.

Gemini 2.5 Pro Preview also addresses critical developer feedback from previous versions. Google notes improved accuracy in function calling and better function trigger rates, enhancing the model’s reliability for complex coding tasks.

Developers currently using the Gemini 2.5 Pro model do not need to make any changes, as the updated version (dated 05-06) has automatically replaced the previous one (03-25). The model remains available through the Gemini App, Vertex AI, and AI Studio at the same pricing.

With Gemini 2.5 Pro Preview, Google sets the stage for a transformative experience at I/O 2025, reinforcing its commitment to empowering developers with state-of-the-art AI tools.