Google has announced that it is launching an experimental version of its Gemini 2.5 Pro neural network to all users of the Gemini app, including free users. This is somewhat unexpected, as the algorithm was only recently introduced as Google’s “smartest AI model,” capable of reasoning in addition to answering user queries.
Features and Availability
Although still experimental, Gemini 2.5 Pro supports various features, including apps/extensions, file uploads, and Canvas. Initially, it was available only to paid Gemini Advanced subscribers. However, over the weekend, Google confirmed that it would be extending access to all Gemini users.
The AI model is now accessible at gemini.google.com and will soon be integrated into the Gemini apps for Android and iOS. Google has emphasized its goal of making its most advanced AI model widely available as soon as possible.
Performance and Future Prospects
Currently, Gemini 2.5 Pro ranks at the top of the LMArena rating, which evaluates models based on user preferences, notes NIX Solutions. It also leads in mathematical (AIME 2025) and scientific (GPQA diamond) tests. Additionally, Google is actively enhancing its neural network’s capabilities for writing program code.
It remains uncertain whether Google will make the stable version of Gemini 2.5 Pro permanently free or if this expansion is part of a broader testing phase. The company has yet to issue official statements on its long-term plans, but we’ll keep you updated as more details emerge.