Google has launched Gemini 2.5, its most advanced AI model yet, designed to tackle complex challenges with enhanced reasoning and coding capabilities, the company announced on its official “The Keyword” blog.
The experimental version, Gemini 2.5 Pro, leads benchmarks significantly and showcases superior performance in various tasks, according to Koray Kavukcuoglu, the CTO of Google DeepMind.
Key Features
Gemini 2.5 models utilize advanced reasoning techniques, allowing them to analyze information, draw logical conclusions, and make informed decisions. The new model excels in creating web applications and transforming code, achieving a notable score of 63.8% on industry-standard coding benchmarks.
With a context window of 1 million tokens (expanding to 2 million), Gemini 2.5 can process vast datasets across text, audio, images, and video.
Gemini 2.5 Pro is now available in Google AI Studio and the Gemini app for Advanced users, with upcoming pricing options for scaled production use. This update marks a significant step towards building AI that can reason and adapt to complex situations, setting a new standard in the field.