Deep Research FAQs

Question 1

What kind of reports does Deep Research generate?

Accepted Answer

Deep Research generates detailed markdown reports that include findings, sources, reliability assessments, and next direction. See exampleout.md in the GitHub Repo.

Question 2

How does Deep Research evaluate source reliability?

Accepted Answer

Deep Research scores sources on a scale of 0-1, providing detailed reasoning for each score. It prioritizes high-reliability sources (≥0.7) and verifies information from less reliable sources.

Question 3

Can I use Deep Research locally?

Accepted Answer

Yes, you can set up a local instance of Firecrawl (a free option) to avoid using an API key. Instructions for setting up a local Firecrawl instance are provided in the README.

Question 4

What is Deep Research?

Accepted Answer

Deep Research is an AI-powered tool that performs deep, iterative research on any topic. It uses search engines, web scraping, and AI to generate comprehensive markdown reports with findings and source reliability assessments.

Question 5

What is MCP and how does Deep Research use it?

Accepted Answer

MCP stands for Model Context Protocol. Deep Research is available as an MCP tool, allowing AI agents to use it for in-depth research capabilities.

Deep Research

About

Key Features

Use Cases