A research agent is an AI system that plans a research task, searches many sources on its own and writes a report with citations.
In plain words
Instead of one search and one answer, the agent breaks the question into smaller research questions. Worker agents search the web or your documents for each part, often at the same time, and summarise what they find. A final step filters and combines those notes into one report that cites its sources. Open-source examples include GPT Researcher and LangChain's Open Deep Research.
A quick example
Anthropic describes a question asking for the board members of every IT company in the S&P 500. Its multi-agent system split the job into tasks for sub-agents and found the answers, while a single agent searching one step at a time did not.
Where you'll meet it
- Literature and market reviews.
- "Deep research" features in AI assistants.
Related
Sources
- Docsgpt-researcher, GPT Researcher (GitHub) · read 29 Sept 2026
- Docsopen_deep_research, LangChain (GitHub) · read 29 Sept 2026
- DocsHow we built our multi-agent research system, Anthropic · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.