After OpenAI and DeepSeek, Google also announces its Thinking model with Gemini

Google has launched its latest Gemini 2.0 Flash Thinking Experimental model, alongside a bunch of other AI models including the Gemini 2.0 Pro Experimental. The new flash thinking model, launched for Gemini app users, now add thinking and reasoning capabilities to Gemini, much like what OpenAI has done with its o1 and DeepSeek's R1 models.

Notably, Google had released its reasoning model in December but didn't get much traction back then. Now, however, the world seems fixated on reasoning models thanks to DeepSeek, which seems to have prompted Google to bundle its model with the Gemini app. In a blog post announcing the launch, Google also said that its Gemini 2.0 Flash model is "currently ranked as the world’s best model."

Just days ago during the Q4 earnings call, Google also announced that the company is expected to spend $75 billion to develop AI infrastructure, as it races to compete with OpenAI, Microsoft, and Meta, as well as a growing presence by low-cost Chinese models.

Google is also rolling out a version of 2.0 Flash Thinking that can interact with apps like YouTube, Search and Google Maps. Google's thinking model works just like you'd expect it to: it shows its thought process, allowing you to see why it responds to queries and prompts in a certain way, as well as trace the model's line of reasoning.