ModelRefs / Research Agent — AI Glossary
Research Agent — AI Glossary
An AI agent that performs multi-step research tasks by searching the web, reading documents, synthesizing findings, and producing reports. Not RAG.
Overview
Research agents (OpenAI Deep Research, Perplexity Research, Gemini Deep Research) issue iterative search queries, read and analyze retrieved pages, identify gaps, and synthesize comprehensive reports. They run for minutes rather than seconds. Evaluated on GAIA and Tau-Bench. Quality depends on the search tool, reading fidelity, and synthesis coherence.
Reference details
| Topic | agents |
|---|---|
| Last reviewed | 2026-06-24 |
Related terms
Commonly confused with
Not RAG. A RAG pipeline retrieves from a corpus you assembled and controls what the model can see; a research agent decides what to look for, chooses sources at runtime, and reads whatever it finds. That is why it works on questions you did not anticipate, and why the untrusted-content problem lands squarely on it — the pages it reads can contain instructions.
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Research Agent — AI Glossary.
Frequently asked questions
What is Research Agent?
An AI agent that performs multi-step research tasks by searching the web, reading documents, synthesizing findings, and producing reports.
What concepts are related to Research Agent?
Closely related concepts include web agent, agentic rag, tool agent.