Announcing Gemini 2.5 Pro (Experimental)
We're excited to announce that Gemini 2.5 Pro (Experimental) is now available on Co! This latest model from Google DeepMind represents a significant leap forward in AI capabilities, particularly in reasoning and complex problem-solving. As Google's most intelligent AI model to date, Gemini 2.5 Pro leads common benchmarks by meaningful margins and showcases exceptional reasoning and coding capabilities.
Key Highlights
- State-of-the-Art Performance: Tops the LMArena leaderboard by a significant margin, indicating superior human-aligned responses
- Advanced Reasoning: Achieves state-of-the-art results across math and science benchmarks like GPQA and AIME 2025
- Enhanced Coding: Excels at creating visually compelling web apps and agentic code applications, scoring 63.8% on SWE-Bench Verified
- Massive Context Window: Processes up to 1 million tokens (2 million coming soon), enabling analysis of vast datasets
- Native Multimodality: Seamlessly handles text, audio, images, video, and entire code repositories
How to Use It
Gemini 2.5 Pro (Experimental) is now available on Co's discovery page. To get started:
- Navigate to the Discover page
- Search for "Gemini 2.5 Pro" and add it to any of your spaces or create a new AI bot using the model by clicking the "Create AI Bot" button
Use Cases
This powerful model excels at:
- Complex reasoning and problem-solving tasks
- Advanced code generation and transformation
- Large-scale document analysis and comprehension
- Multimodal content understanding and generation
- Scientific and mathematical problem-solving
With its combination of enhanced reasoning capabilities, massive context window, and superior performance across benchmarks, Gemini 2.5 Pro enables developers to tackle increasingly complex AI challenges with unprecedented accuracy and efficiency.