Google has introduced a new experimental AI model, Gemini 2.0 Flash Thinking Experimental, aimed at tackling complex problems in programming, mathematics, and physics.
The model, now available on Google’s AI Studio, is designed to excel in multimodal understanding, reasoning, and coding. It pauses before responding to prompts, analyzing and explaining its reasoning to improve accuracy.
Built on the Gemini 2.0 Flash platform, this model represents Google’s entry into the growing field of reasoning-focused AI.
Unlike traditional models, reasoning AI fact-checks itself, which helps avoid common errors but requires more computational power and time to deliver solutions.
Despite its promise, early tests have revealed flaws, such as struggling with basic tasks like counting letters in a word.
Google joins competitors like DeepSeek and Alibaba, who are also exploring reasoning models. However, these models face criticism for being resource-intensive and slower than conventional AI systems.
While still in its early stages, Gemini 2.0 signals Google’s commitment to advancing AI through innovative approaches to generative reasoning.
