In an unexpected move, Google announced on Saturday that it is offering the latest 2.5 Pro (experimental) model to all free Gemini app users.
Gemini 2.5 Pro was announced on Tuesday as Google’s “most intelligent AI model,” which integrates thinking capabilities instead of presenting it as a separate version. It is currently in the “experimental” phase, but supports various features, including app/extensions, file uploads, and (to date) Canvas.
At the time of launch, it was only available to Gemini Advanced customers, with Google One AI Premium costing $19.99 per month (in the US).
On Saturday afternoon, Google announced that it had “decided to offer Gemini 2.5 Pro (experimental) to all Gemini users starting today.”
It is already widely available on gemini.google.com and will soon be available on the Android and iOS apps. Both platforms have recently improved the feature to better remember which model you selected.
Google says it wants to “get its most intelligent model into the hands of as many people as soon as possible.” Gemini 2.5 Pro (experimental) is currently topping the LMArena leaderboard, leading in math and science benchmarks. Google is also working on improving coding performance.
Looking ahead, it is unclear whether Google will make the stable version of 2.5 Pro available for free to all Gemini users or if today’s move is just for the testing phase. Earlier, it seemed that 2.0 Flash Thinking and 2.0 Flash would be offered for free.
That said, in recent months, Google has been making more and more features available to all users, including Deep Research, Gems, saved information, document uploads/analysis, and the ability to create images of people. For advanced customers, the main feature is the 1 million token reference window, along with higher usage limits.
