Deep research (AI)
Deep research is an AI agent feature that plans a research question, runs many web searches, reads the sources it finds and writes a long report with citations, taking minutes rather than seconds.
Also called: research agent
By Divya Sharma · Checked Nov 15, 2026 · How we review
A normal chat answer comes from one pass of the model, perhaps with a single search. Deep research turns the same question into a small project. The agent writes a plan, searches, reads pages, notices gaps, searches again, and finally writes a structured report with links to what it read.
How it works
Under the hood it is an agent loop with search and page-reading tools. Perplexity describes its version as doing dozens of searches and reading hundreds of sources before it writes, and says most tasks finish in a few minutes. ChatGPT, Gemini and Perplexity all offer a mode under this name, and open-source projects such as GPT Researcher do the same job with your own model and search keys.
Deep research versus AI search
An AI search engine answers in seconds from a handful of results. Deep research spends far more time and tokens to cover a topic in breadth, which suits literature reviews, market scans and due diligence rather than quick facts.
Limits
More sources do not guarantee a correct report. The agent can lean on weak pages, miss paywalled material, or state a hallucination with a citation that does not support it. Treat the report as a draft: open the key sources and check that they say what the report claims.
Related terms
- AI agent
- An AI agent is a system in which a language model decides its own next steps: it calls tools, reads the results and repeats until the task is done, instead of following a fixed script.
- AI search engine
- An AI search engine is a search service that runs web searches for a query, reads the results with a language model and returns a written answer with links to its sources, instead of only a list of links.
- Agent loop
- An agent loop is the repeating cycle at the heart of an AI agent: the model asks for a tool call, the application runs it and returns the result, and the model decides what to do next.
- Grounding
- Grounding is tying a language model's answer to specific, checkable sources, such as search results, documents or a database, so that its claims rest on supplied evidence rather than only on training data.
- Hallucination (AI)
- A hallucination is output from an AI model that sounds fluent and confident but is false, unsupported by its sources or made up, such as an invented citation, quote, statistic or API function.