There was the time when Large Language Models just blurt out whatever they have in mind, based on what they learned.
But since ChatGPT from OpenAI took off and sparked an increasingly tense competition, players in this emerging industry began to realize how unreliable these AI chatbots are in, at least in term of accuracy.
To address this, they started developing ways to make the reason, and question their own thoughts, reducing hallucination
They also began creating AIs that are capable of doing more thorough research, using whatever they have in their disposal to scout the web and reason, and combine all data into one comprehensive report.
Following ChatGPT's deep research feature and Google's enhanced Gemini 2.0, as well as the emergence of DeepSeek that worries the U.S., Perplexity AI followed by introducing its own "deep research" feature.
The feature is designed to give users the tools they need to do extensive research in just minutes, providing comprehensive answers to their complex queries.
Introducing Deep Research on Perplexity.
Deep Research lets you generate in-depth research reports on any topic.
Available to everyone for free—up to 5 queries per day for non-subscribers and 500 queries per day for Pro users. pic.twitter.com/obovx7YEUF— Perplexity (@perplexity_ai) February 14, 2025
In a blog post, Perplexity said that:
The Deep Research feature on Perplexity, is similar those from others, in which it allows users to have the AI conduct more thorough and in-depth investigations into a topic.
What this usually means, the AI will initiate a more enhanced web searches to find most recent and relevant data, rather than relying only on its pre-existing knowledge; doing multi-step reasoning to break down complex queries; document analysis; fact-checking sources for validation, and more.
All that in order to create get more nuanced, detailed, and expert-level explanations rather than just surface-level answers.
In short, it’s like having an AI-powered research assistant that can dig deeper, analyze sources, and provide more refined insights instead of just giving quick, general responses.
Read: Meet 'Perplexity Assistant,' An AI Assistant That Can Reason, Search, Use Apps, And More
Perplexity Deep Research combines advanced web search, reasoning, and coding to iteratively search and refine analysis into clear, comprehensive reports. pic.twitter.com/iQLOSwA87q
— Perplexity (@perplexity_ai) February 14, 2025
But what it does best, is leveraging what Perplexity already excels at in answering questions.
The AI from Perplexity was born as a conversational search engine that uses LLMs to "reinvent" web searches by answering queries using sources from the web and cites links within the text response.
Building on top of that, Perplexity enhances the AI with Deep Research, by making it able to answer next-level questions by spending only 2-4 minutes doing the work of what a human expert can take several hours to perform.
As explained by Perplexity, the Deep Research feature is equipped with search and coding capabilities.
It can do searches, reads documents, and reasons about what to do next, refining its research plan as it learns more about the subject areas.
"This is similar to how a human might research a new topic, refining one’s understanding throughout the process," Perplexity explains.
After spending a few moment to compile everything it needs to tell, the AI will then create "a clear and comprehensive report," where source materials are presented after being evaluated.
Users can then export and share the report to a PDF or document, or convert it into a Perplexity Page and share it with colleagues or friends.
We believe everyone deserves powerful research tools.
That's why we're making Perplexity Deep Research free for all.
Try it now by selecting "Deep Research" from the mode selector in the search box.
Read the full blog article here: https://t.co/U8DLE2oLGx— Perplexity (@perplexity_ai) February 14, 2025
As for how good this AI is, Perplexity sad that its Deep Research feature allows its LLM to surpass many others in the competition—scoring 93.9% accuracy on the SimpleQA benchmark.
The company said that Deep Research on Perplexity scores 21.1% on Humanity’s Last Exam, outperforming Google Gemini Flash Thinking, OpenAI o3-mini, OpenAI o1, DeepSeek-R1, and some others.
Perplexity has made this Deep Research feature available and accessible to all users.
However, free users and paid users have different perks.
Those who use Perplexity with free accounts are allotted to only up to five queries per day, while Pro users benefit from an expanded limit of 500 queries daily.
Initially only available on the web, Perplexity AI plans to extend this feature to iOS, Android, and Mac platforms.
To utilize this Deep Research feature, users should go to the Perplexity AI website and select 'Deep Research' from the mode options in the search box before submitting their query. This streamlined process enables users to generate in-depth research reports on any topic efficiently.
Deep Research is now a commodity thanks to Perplexity pic.twitter.com/Fk8yvPTLzV
— Aravind Srinivas (@AravSrinivas) February 14, 2025
Further reading: Perplexity AI Allegedly Disobey A Web's Standard, And Scrape Content In Places It Shouldn't