Unlocking Insights: Google Integrates Gemini 2.5 Pro Experimental into Enhanced Deep Research Tool for Advanced Users

Google is broadening the capabilities of its Gemini 2.5 Pro Experimental artificial intelligence (AI) model by integrating it with additional tools. On Tuesday, the tech firm based in Mountain View revealed that its newest Gemini variant will now enhance the agentic Deep Research functionality. This feature is currently exclusive to paid subscribers of the AI platform, with free users continuing to utilize the AI agent powered by the Gemini 2.0 Flash Thinking (experimental) model. Google asserts that the latest large language model (LLM) will enable users to experience significant enhancements in the tool’s analytical reasoning abilities.

Gemini 2.5 Pro to Enhance Deep Research

In a blog entry, the tech company announced the integration of its latest foundational model with Deep Research. The Gemini 2.5 Pro Experimental AI model was unveiled last month, featuring numerous improvements. This marks the first series of Gemini models to include integrated reasoning capabilities, making all Gemini 2.5 AI models inherently “Thinking” models. Consequently, incorporating this model into Deep Research is a fluid process.

Google reported that it conducted an internal evaluation of Deep Research utilizing Gemini 2.5 Pro Experimental, and the evaluators favored the generated reports compared to those from competitors. Specifically, the company noted that the testers observed a remarkable enhancement in analytical reasoning, synthesis of information, and the crafting of insightful research reports.

At present, Deep Research with Gemini 2.5 Pro Experimental is exclusively accessible to Gemini Advanced subscribers. Users with this subscription can engage with the AI agent via web platforms as well as the Android and iOS applications. In the meantime, users on the free version of Gemini may continue to leverage Deep Research with the Gemini 2.0 Flash Thinking (experimental) model.

It is important to note that Google initially launched its agentic Deep Research in December 2024, originally limiting access to Gemini Advanced users. Later, in 2025, the AI agent was made available to all users.

Deep Research has the capability to develop multi-step research strategies, perform web searches, and gather information on specific subjects and related domains. Following this, it analyzes the collected data, prepares a comprehensive report, and presents the findings to the user in a detailed format.