Edgar Cervantes / Android Authority
TL;DR
- Google has reportedly brought its Deep Research tool to the Gemini app on Android.
- This tool, which was previously restricted to the web, allows Gemini to create detailed reports on your behalf.
- This also comes a few days after OpenAI launched its own deep research feature.
Google released its Deep Research tool for Gemini late last year, helping you get in-depth research done in minutes. The tool was only available on the web, but there’s good news if you want to use this feature on the go.
9to5Google reports that Deep Research is now widely available in the Gemini app for Android. Only one Android Authority team member reported seeing the feature on their phone, suggesting a gradual release.
This tool, which is restricted to Gemini Advanced users, is identical to the web-based version. That means you have to first select the “1.5 Pro with Deep Research” option from the drop-down menu up top. From here, you can enter your query, and the tool will show you an outline of the report it will create. You can either edit this outline/plan or tap “start research” to have Google sift through the web and build the report. This report will also offer sources, as you’d expect.
Google notes on its Deep Research support page that the report takes five to ten minutes to create, but that more complex reports can take longer. The company adds that you can leave the chat and then get a notification when the report is ready, and this is reflected in the Gemini app too.
Colleague Mitja Rutnik tried Deep Research at the time of its release and noted that he had “mixed feelings” about it. Mitja found that the tool fared better with broader and simpler topics like interior design and AI chatbots. However, he called it “useless” if you wanted to generate something like a stock market report for a given day.
Google isn’t the only company offering a tool like this. OpenAI followed in the search giant’s footsteps earlier this week by launching its own Deep Research feature. This is powered by the OpenAI o3 model and promises to generate a comprehensive report within 30 minutes.
