Skip to main content

Using Research on Claude.ai

Updated today

Research is a beta feature available to individuals in the United States, Japan, and Brazil who are using Claude.ai on the web or desktop apps and who are on Max, Team, or Enterprise plans.

Research transforms how Claude finds and analyzes information. Claude operates agentically, conducting multiple searches that build on each other while determining exactly what to investigate next. It explores different angles of your question automatically and works through open questions systematically.

With Research, Claude delivers thorough answers in minutes, complete with easy-to-check citations so you can trust Claude's findings.

Note: You must have web search turned on for Research to function.

Enabling/Disabling Research

To enable:

  1. Look for the Research dropdown in your chat interface

  2. Select Research from the picker

To disable:

  1. Look for the Research dropdown in your chat interface

  2. Select Default from the picker

Using Research

Once Research is turned on, you can ask Claude a question. Claude will kick off the Research process across your internal context (such as Gmail, Google Calendar and Google Docs when connected) and the web.

FAQs

Q: I have the Research chat toggle on, but Claude doesn’t seem to be researching. What should I do?

A: In this case, you can steer and prompt Claude to use Research by saying “Claude, please use the Research tool to…”

Q: I have the Research chat toggle on, and Research should be pulling against my internal documentation via the Google integrations I have turned on, but isn’t. What should I do?

A: In this case, you should steer and prompt Claude to “Pull relevant context from [relevant internal knowledge source].”

Q: Does Research count against message limits differently than normal conversations?

A: Research is subject to the same limits as standard Claude conversations. However, Research sessions can use up your limits faster due to Claude retrieving multiple sources and providing comprehensive responses.

Did this answer your question?