Google Launches Gemini 2.5 Pro and Budget-Friendly Flash-Lite Model

Google is ramping up its AI game with the official release of the stable versions of Gemini 2.5 Pro and Flash, alongside the introduction of the new wallet-friendly Flash-Lite model. Developers now have greater control over computing resources, and those seeking high output at minimal cost can opt for a solution that expenses just a fraction of previous offerings.

Gemini 2.5 Flash was finalized shortly after the Google I/O conference, but as of June 5, the Gemini 2.5 Pro model is also fully ready, with all previous issues addressed in this latest build. Both models are accessible via Google AI Studio and the Vertex AI platforms.

The most intriguing update is Flash-Lite. This new model handles text, images, and video processing at up to three times lower cost than standard Flash, and generates responses at up to six times lower cost. While less powerful than Flash, Flash-Lite is ideal for those who pay per token, making every cent count.

Flash and Flash-Lite are also being integrated into Google Search, powering the new AI mode and AI overviews. For complex queries, Gemini 2.5 Pro will be activated, while simpler tasks will be handled by Flash or Flash-Lite.

For end users, there are no major changes expected in the Gemini app, as these models already operate behind the scenes. Notably, the "Preview" label is being removed from the 2.5 Pro model. Free users will have limited access to Pro, while subscribers get more: those on the Pro plan can use it up to 100 times per day, and AI Ultra subscribers enjoy the broadest access.