Google’s reasoning artificial intelligence was introduced + video

Google has introduced a test version of a new artificial intelligence model that has the ability to “reasoning» and will compete with OpenAI’s o1. Google modelGemini 2.0 Flash Thinking” and can answer your complicated questions.

In the Google blog, it is written about this artificial intelligence model:

“Gemini 2.0 Flash Thinking Mode is an experimental model that is trained to create a ‘thinking process’ that it goes through as part of its response. As a result, Thinking Mode has better reasoning ability in its answers than the base model of Gemina 2.0 Flash.

“Sundar Pichai”, the CEO of Google, also mentioned Gemina 2.0 Flash Thinking Mode as “the most thoughtful model ever made by Google” on X social network.

How Google Reasoning Artificial Intelligence works

In a post on X, Google DeepMind Senior Scientist Jeff Dean explained that the model was “trained to think to reinforce its reasoning” and was as fast as Gemini Flash 2.0.

In the video she shared below, Gemina 2.0 Flash Thinking helps solve a physics problem by “thinking”.

According to his description, the new model supports 32,000 input tokens (about 50-60 pages of text) and can generate 8,000 tokens per response as output. Google says of its new model that it is the best choice for “multimodal understanding, reasoning and coding”.

In another example, demonstrated by Google product manager Logan Kilpatrick, the company’s reasoning AI solves a challenging puzzle that includes visual and textual clues.

You can now try Gemini 2.0 Flash Thinking in Google AI Studio.