Google has unveiled its latest AI model, Gemini 2.0 Flash Thinking, designed for complex problem-solving with enhanced reasoning capabilities.
According to CEO Sundar Pichai, this is "our most thoughtful model yet." The new model supports 32,000 tokens of input and can generate responses of up to 8,000 tokens.
Gemini 2.0 Flash Thinking allows users to view its step-by-step reasoning, addressing concerns about AI's transparency. This feature sets it apart from OpenAI's models, which do not provide such insights.
The model's multimodal capabilities enable it to process various data types, including images and text, making it versatile for different applications. It is currently available for testing on Google AI Studio.
Independent analysis from LM Arena has recognized Gemini 2.0 Flash Thinking as the top-performing model in its category, marking a significant advancement in the AI landscape.