Google's new experimental Gemini 2.5 model rolls out to free users
Briefly

Google's Gemini 2.5 Pro (Experimental) model has been made accessible to free users following its initial launch for subscribers. This model includes enhanced simulated reasoning capabilities, aiming to minimize past AI mistakes. Although it performs impressively well, gaining the top rank in the LMSYS Chatbot leaderboard, free users will not have access to all features available in the paid tier, such as file uploads. Limitations on context length for the free version remain unspecified, but it's noted that users can input several thousand words without issues.
Google's swift rollout of Gemini 2.5 Pro (Experimental) provides free users with access to advanced AI features, although not all functions of the paid version.
Gemini 2.5 incorporates simulated reasoning, potentially reducing errors commonly seen in AI outputs and demonstrates high performance in the LMSYS Chatbot leaderboard.
The free version of Gemini 2.5 Pro is accessible online and will soon be available in the mobile app, enhancing usability for non-subscribers.
While the free tier allows experimentation with new AI capabilities, it restricts features like file uploads and operates with undisclosed limitations.
Read at Ars Technica
[
|
]