Perplexity takes on ChatGPT and Gemini with its own Deep Research Feature
Back in December, Google's Gemini introduced a new feature called Deep Research that can delve into complex subjects and create reports with relevant citations. Earlier this month, OpenAI introduced a Deep Research feature in ChatGPT that uses the o3 model's reasoning capabilities to complete complex tasks. Today, Perplexity joined Gemini and ChatGPT by introducing its own Deep Research feature that can do in-depth research and analysis for users.
How Perplexity's Deep Research Feature Works
Perplexity's Deep Research feature works similarly to the other existing Deep Research features. It performs a web search based on the query, reads those search results, reasons through the content, and prepares a report for users. It completes all these steps for most research tasks in under 3 minutes.
OpenAI and Google have made their Deep Research feature available for both free and paid tier users. Perplexity is following the same model for its Deep Research feature by making it free for all its users. While Perplexity Pro users will get unlimited Deep Research queries, free users will only get a limited number of queries per day.
According to Perplexity, its Deep Research feature scored 20.5% accuracy on Humanity’s Last Exam, significantly higher than Gemini Thinking, o3-mini, o1, and DeepSeek-R1. It is important to note that OpenAI's Deep Research feature is still ahead of Perplexity and others.
The Deep Research feature is now available on the Perplexity web experience, and it is coming soon to iOS, Android, and Mac. Perplexity web users can select “Deep Research” from the mode selector in the search box to use this feature.
