Google has added its experimental artificial intelligence model to the Gemini app, capable of «reasoning» and explaining how it answers complex questions. The Gemini 2.0 Flash Thinking update is part of a large-scale rollout of Gemini 2.0 AI. The new products also include the flagship Gemini 2.0 Pro model.
Gemini 2.0 Flash Thinking will be available in a drop-down list of models on PC and in the mobile app. In addition, Google is introducing another version of artificial intelligence that will be able to interact with YouTube, Google Search, and Google Maps. It was introduced in December 2024 and is expected to compete with other so-called «thinking» AI models, such as o1 from OpenAI and R1 from DeepSeek.
These models work by breaking down problems into smaller, manageable steps, allowing them to «think» about queries before coming up with solutions. The expected outcome is to achieve better and more accurate results, but this often comes at the cost of increased time to achieve them.
Google has also released an experimental version of Gemini 2.0 Pro. The successor to Gemini 1.5 Pro is supposed to provide «better factuality» and «better performance» for coding and math-related tasks. Gemini 2.0 Pro is described as Google’s «most capable model to date and will be available to Advanced Gemini app users and people with access to Vertex AI and AI Studio.
Gemini 2.0 Flash — the latest version of Google’s high-performance AI model — is also now publicly available to developers in AI Studio and Vertex AI, following its rollout to Gemini web and mobile apps last week.
At the same time, Google has introduced a new model called Gemini 2.0 Flash Flash-Lite, which matches 1.5 Flash in terms of speed and price, but surpasses it in most» criteria. Gemini 2.0 Flash-Lite is being launched in public preview in Google AI Studio and Vertex AI.
Source: The Verge
Spelling error report
The following text will be sent to our editors: