Kotaemon
Lets you ask questions about your own documents and see cited answers in a web page
by Cinnamon
GPT Researcher
by Assaf ElovicAgentResearch something
Turns a research question into a cited report from web pages or your own documents
About GPT Researcher
GPT Researcher is an open-source agent that researches a question on the web or in your own files and writes a report with its sources. Its stated mission: accurate, factual information for individuals and organizations.
For IndividualsOrganizations
Made by Assaf Elovic
GPT Researcher is an open-source project published on GitHub by Assaf Elovic, with a website at gptr.dev.
- Stars
- ★ 31k
- License
- Apache-2.0
- Cost
- Usage-based
- Platforms
- Mac · Win · Linux
- Level
- Advanced
Overview
GPT Researcher is an open-source research agent. You give it a question; it plans sub-questions, gathers pages from the web or files from your own folder, and writes a long report that lists its sources, which you can export to PDF or Word. It runs on your computer through Docker or Python and calls the AI model and search service whose keys you provide. It suits people who want a sourced first draft on a topic and can follow a command-line setup.
Its deep research mode explores subtopics in breadth and depth, and the README puts a run at about five minutes.
Try these first
Start with the first one: it is a small task you can check easily.
- Research [a narrow topic you know well] and write a short report that lists every source it used.First try · paid API use
- Summarize what the documents in my folder say about [question].needs DOC_PATH
- Run a deep research on [topic].paid API use
- Write the report on [topic] again and export it as a Word file.writes files
Example prompts written by Agent Flows Hub.
Features
From the creator's documentation.
| Feature | What it does | Runs |
|---|---|---|
| Cited reports | Plans research questions, gathers sources and writes a report that lists them | Web and AI model API |
| Local documents | Researches PDF, text, CSV, Excel, Markdown, PowerPoint and Word files in a folder you choose | Your files, AI model API |
| Deep research | Explores subtopics in a tree; the README says about 5 minutes and about $0.4 per run in its example | AI model API |
| Export | Saves reports as PDF, Word and other formats | Your computer |
| Web interface | A lightweight page served by the app, or a NextJS app, opened in your browser | Your computer |
Good to know
You need a sourced overview of a topic, such as a market, a technology or a news event. GPT Researcher breaks the question into parts, reads web pages or your own files for each part, and assembles one report that lists where each piece came from.
Limits
- It calls an AI model and a search service with your own API keys, so each run is billed by those providers.
- The README setup uses git, Docker or Python 3.12 or newer, and environment variables; it names no one-click installer.
Notes by Agent Flows Hub (original project name: GPT Researcher), written from the creator's documentation, not from running it ourselves. GPT Researcher: README and setup · GPT Researcher: documentation
Stuck setting it up?
Tell us what happened. We take a look first; if it needs paid work, you get an estimate before anything starts.
More to research something
Other agents
- grok-cli Hands coding tasks to a terminal agent that runs on xAI's Grok models
- HanaAgent A desktop AI assistant with memory that works on your files and runs scheduled tasks
- Hermes Agent Gives you an AI assistant that remembers you and does tasks from Telegram or Slack
- InkOS Helps you plan, write and revise novels, short stories and scripts with an AI agent
- jcode Runs many coding agents side by side in your terminal, using little memory per session
- Kilo Code Hands your coding tasks to an agent in VS Code, JetBrains or the terminal