Kotaemon

by CinnamonAppResearch something

Lets you ask questions about your own documents and see cited answers in a web page

View on GitHub

The problem it solves

You have a stack of PDFs and need one answer from them, but a chatbot rarely tells you which page it came from.

For End users doing QA on documentsDevelopers building a RAG pipeline

Before: You open each file, search for keywords and hope you found the right page.
After: You ask in plain words and get an answer with citations that open at the highlighted passage in a PDF viewer.

Stars: ★ 26k
License: Apache-2.0
Cost: Usage-based
Platforms: Mac · Win · Linux
Runs: Local · Self-hosted

Screenshots

From the project's own repository.

The Chat tab of the Kotaemon web interface — Image: Cinnamon (Apache-2.0). Original file

Overview

Kotaemon is an open-source web app for chatting with your documents. You upload files such as PDFs, connect an AI model from a provider like OpenAI or one running on your own computer through Ollama, and ask questions. Answers come with citations that open in a built-in PDF viewer with highlights. It supports several users, with private and shared file collections.

By default it gives detailed citations, and you can open them in a built-in PDF viewer with highlights.

Try these first

Start with the first one: it is a small task you can check easily.

What does [uploaded PDF] say about [topic]? Show the page it comes from.First try · text sent to model
Compare what [file A] and [file B] say about [topic].text sent to model
List the figures and tables in [uploaded PDF] that mention [topic].text sent to model
Break this question into steps and answer it from my files: [question]text sent to model

Example prompts written by Agent Flows Hub.

Features

From the creator's documentation.

Feature	What it does	Runs
Cited answers	Answers come with citations that open in an in-browser PDF viewer with highlights	Your computer
Model choice	Use API providers such as OpenAI, Azure, Cohere or Groq, or local models through Ollama	Your model provider
Hybrid search	Combines full-text and vector search with re-ranking to find passages	Your computer
Figures and tables	Answers questions about documents with figures and tables, with selectable parsing options	Your computer
Several users	Multi-user login with private and public file collections you can share	Your computer
Settings page	Adjust retrieval and generation settings, including prompts, in the page	Your computer

Good to know

You have PDFs or other documents and want answers from them in plain words. Kotaemon indexes your files, answers with citations, and opens the cited passage in a PDF viewer so you can check it.

Limits

The app needs access to an AI model, added in the Resources tab (with an embedding model) or set in the .env file.
The default login is admin / admin, and the docs say to change it right after the first sign-in.
Files other than PDF, HTML, MHTML and XLSX need the optional Unstructured package.

Notes by Agent Flows Hub (original project name: kotaemon), written from the creator's documentation, not from running it ourselves. Kotaemon README · Kotaemon getting started guide

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.

Get help

How to start

Install itPaste the install request into an AI agent you already use, or run the official command yourself or follow the official guide.macOS, Windows, LinuxDownload the installermacOS, Linuxdocker run \ -e GRADIO_SERVER_NAME=0.0.0.0 \ -e GRADIO_SERVER_PORT=7860 \ -v ./ktem_app_data:/app/ktem_app_data \ -p 7860:7860 -it --rm \ ghcr.io/cinnamon/kotaemon:main-full
Open itFollow the first-run steps in the app.
Give it a first taskUpload one short PDF you know well, ask a question whose answer is on a page you can find, then open the citation and compare it with that page.

Steps follow the creator's instructions. The request tells your agent to read them first and wait for your OK.

Before you start

A computer with macOS, Windows or Linux.
An API key from a model provider like OpenAI, or a local model through Ollama
About 20 minutes for the offline installer: unzip it and run one script
Pay per use. The AI service it calls charges for what you use.

Details

Made by: Cinnamon
License: Apache-2.0 Terms
Platforms: macOS, Windows, Linux
Runs on: Your computer, Self-hosted server

View source Official docs Report a problem

For AI agents

This page as a Markdown install brief:

/en/library/kotaemon/install.md

Other apps

LangBot Puts an AI chatbot into WeChat, QQ, Lark, Telegram, Slack, Discord and other chat apps
Langflow Lets you build AI agents and workflows by connecting blocks on a visual canvas
LearnHarness Teaches you a hard topic step by step, starting from what you already understand
LibreChat Gives you a self-hosted ChatGPT-style app where you switch between many AI models
LM Studio Finds, downloads and runs open AI models on your computer in one desktop app
LobeHub Lets you build a team of AI agents that chat, write and run scheduled tasks for you

See all 122 apps · When a setup fails

GitHub