Gemini Advanced subscribers now have access to Deep Research powered by Gemini 2.5 Pro Experimental, which is recognized as the most capable AI model in the world according to reasoning benchmarks and Chatbot Arena comparisons.

Gemini Deep Research acts as a personal AI research assistant. With the latest model, its effectiveness has increased across all phases of research. According to internal testing, users preferred the research reports generated by Gemini Deep Research using 2.5 Pro over competing deep research tools by a margin of more than two to one.

This updated model has demonstrated significant improvement in analytical reasoning, information synthesis, and the creation of insightful reports. Testers have noted enhanced capability in producing comprehensive, easy-to-understand content across a wide variety of topics.

Gemini Advanced subscribers can now use Deep Research across the web, Android, and iOS platforms. The feature enables users to create in-depth reports efficiently, saving considerable time during the research process.

A notable addition to the offering is the Audio Overviews feature. This allows users to convert research reports into podcast-style audio, making it possible to consume content on the go.

To access the new feature, users can select "Gemini 2.5 Pro (experimental)" from the dropdown menu and then tap "Deep Research" in the prompt bar. Additional information is available on Gemini's official website.

PHOTO: ANTARA/GOOGLE DEEPMIND

This article was created with AI assistance.