Google has announced the launch of Gemini 2.5 Pro, its latest AI model with enhanced reasoning capabilities. This version can break down tasks logically before providing answers, significantly improving accuracy and performance. Initially, access was limited to paying users, but it is now free for all. The CTO of Google DeepMind details how this model incorporates contextual analysis and nuanced decision-making. Gemini 2.5 Pro builds on earlier versions, having achieved top scores in benchmark rankings and showcasing advanced performance for complex tasks in AI.
Gemini 2.5 Pro is now selectable as one of the LLMs within the Gemini app and Google AI Studio, previously accessible only to Advanced users.
Gemini 2.5 Pro has 'think' capabilities, allowing it to split tasks into multiple steps for better performance and accuracy.
According to Koray Kavukcuoglu, this reasoning ability includes analyzing information, drawing logical conclusions, and making substantiated decisions.
Google’s Gemini 2.5 Pro has outperformed its predecessor in LMArena ranking and is deemed the most advanced model for complex tasks.
Collection
[
|
...
]