Google Expands Gemini 2.0 Rollout With Flash Thinking, Agentic and Pro Models

Google announced the availability of several Gemini 2.0 artificial intelligence (AI) models on Wednesday. The Mountain View-based tech giant is rolling out its Gemini 2.0 Flash Thinking Experimental AI model to mobile apps and the web client. Alongside this, an agentic version of the AI model is also being rolled out that can interact with certain apps. The company is also releasing an experimental version of Gemini 2.0 Pro to paid subscribers. Additionally, a lite version of the 2.0 Flash is also being rolled out as a public preview.

Google Rolling Out Multiple Gemini 2.0 Series AI models

In a blog post, the tech giant detailed all the different models currently being released to users. Some of these are available to the free users of Gemini, some only to the paid subscribers, and others are exclusively available to developers.

The most notable among these is the Gemini 2.0 Flash Thinking, a reasoning-focused model comparable to the DeepSeek-R1 and OpenAI’s o1 models. It was first released in December 2024, but so far it could only be accessed from Google’s AI Studio.

Now, the company is rolling out the model to all Gemini app and website users. The new AI model will be available via the model selector option at the top of the interface. Gadgets 360 staff members have not seen the model yet, however, it should be available globally in the coming days. Notably, it is unclear whether the free tier will face any rate limits for using the Thinking model.

Alongside, the tech giant is also making an agentic version of the 2.0 Flash Thinking available to users. This model can interact with apps such as YouTube, Google Search, and Google Maps. With this integration, users should be able to ask Gemini to complete certain tasks on these apps. The extent of its ability is currently not known.

For Gemini Advanced users, Google is rolling out an experimental version of Gemini 2.0 Pro, the high-performance frontier model of the 2.0 series. The model is said to be good at breaking down complex problems as well as in coding and mathematics-related tasks. This is the tech giant’s most advanced model, with a context window of 2 million tokens. The application programming interface (API) of the model will also be able to call tools such as Google Search and code execution. It will also be available in Google AI Studio and Vertex AI.

Developers will now be able to access Gemini 2.0 Flash-Lite. The company said it offers better performance than the 1.5 Flash while retaining the speed and cost. It has a context window of one million tokens and accepts multimodal input. In addition, Google is also rolling out the 2.0 Flash model to developers via the Gemini API. Currently, it can handle text-based tasks, and the company will add image generation and text-to-speech capabilities in the future.

Both of these models will also be available in the Google AI Studio and Vertex AI.