Gemini 2.5 is the latest AI model designed to tackle complex problems with enhanced reasoning and advanced coding capabilities. The experimental version, Gemini 2.5 Pro, leads on common benchmarks and showcases strong reasoning and code skills, ranking #1 on LMArena. This model can reason through its thoughts before responding, providing improved accuracy and performance. Gemini 2.5 Pro excels in tasks requiring advanced reasoning and coding, scoring high on industry benchmarks like GPQA and SWE-Bench Verified. With a million token context window and native multimodality, Gemini 2.5 Pro is setting a new standard in AI capabilities. Get ready to experience the future of AI with Gemini 2.5 Pro.
https://blog.google/technology/google-deepmind/gemini-model-thinking-updates-march-2025/