Google is updating the gEMINI for the iPhone With an important novelty that has been available in the application for Android, as has revealed 9to5google. This is the function of Deep Research. This offers a crucial advantage for both models and chatgpt, and for Siri, making the Apple virtual assistant pass to a very background.
Specifically, the Deep Research function, which now comes to Gemini for the iPhone, “I use advanced reasoning and long context capabilities to act as a research assistant”. That is, it allows users to ask questions so that the chatbot throws complete results, with many more data from the usual and, even, with reports that corroborate that data.
With the Deep Research function, Gemini It takes something more (between 5 and 10 minutes) in throwing the results that with the normal version of the AI. This is because the model seeks the content on the Internet, checks its veracity, corroborates it with other data, compares reports and examines the content to offer a response as precise and complete as possible. In fact, when the user uses the Deep Research mode, Gemini shows a search plan that the user can modify or accept.
Deep Research for Gemini app on the iPhone and its limitations
At the moment, Deep Research is available with the model Gemini 1.5 Probut it is likely that Google later implements it in Gemini 2.0, version of the model that is currently in the testing phase. The feature, of course, is only available for Gemini Advance users, the service under the subscription of Google AI that is priceless of 21.99 euros per month.
In addition, Deep Research’s function is somewhat limited. That is to say, Google establishes a maximum number of deep searches per day, and it is not possible to make different consultations simultaneously. The app will warn the user when it has reached the limit.
To use Deep Research in Gemini, Simply update the app for the iPhone to the latest version. Once inside the chatbot, it is necessary to press on the upper screen tab and select Gemini 1.5 Deep Research. Users can easily change among the different versions of the model if they do not want to make deeper searches.
