Recently announced its most powerfulArtificial Intelligence Model Gemini 2.0After first providing a lightweight Flash version for testing, Google earlier announced an artificial intelligence model called Gemini 2.0 Flash Thinking, which claims to be able to answer more logically complex questions through "thinking".
According to Jeff Dean, chief scientist of Google DeepMind, Gemini 2.0 Flash Thinking mainly adds the "thinking" function, which is somewhat similar in form.OpenAI's o1 model, but still maintains the agile execution speed of the Flash version.
However, while it's called "thinking," it's actually more like breaking down a larger question into smaller ones, and then answering each one to arrive at the answer. For example, when Gemini 2.0 Flash Thinking is asked to solve a complex math problem, it breaks it down into its various calculations and calculates each one individually to arrive at the final answer.
Gemini 2.0 Flash Thinking also has multimodal input capabilities, which means it can process images and text content simultaneously. Just like solving complex problems, it will break down the problem one by one and find the solution.
Currently, Gemini 2.0 Flash Thinking will be available for testing through the Google AI Studio page.
Introducing Gemini 2.0 Flash Thinking, an experimental model that explicitly shows its thoughts.
Built on 2.0 Flash's speed and performance, this model is trained to use thoughts to strengthen its reasoning.
And we see promising results when we increase inference time…
— Jeff Dean (@JeffDean) December 19, 2024
