MaxKB

by FIT2CLOUDAppGet organised

Builds a chat assistant that answers questions from your own documents, on your server

View on GitHub

The problem it solves

Customers keep asking questions your manuals already answer, and your team spends the day pasting the same replies.

For Customer service teamsInternal help desksSchools and research groups

Before: Staff search manuals and shared folders by hand, then type out the same answers again and again.
After: A chat assistant built from your documents answers on a shareable chat page or in a widget on your website.

Stars: ★ 23k
License: GPL-3.0
Cost: Free
Platforms: Mac · Win · Linux
Runs: Self-hosted

Overview

MaxKB is an open source platform from FIT2CLOUD for building AI assistants that answer from a knowledge base. You upload documents or add web pages, connect a cloud or local model, and publish a chat app you can share or embed on your website. Visual workflows and tools let the assistant take further steps beyond answering. It suits teams that want a customer service or internal help desk assistant running on their own server.

It turns a set of company documents into a chat assistant through a few steps in the browser, and it runs on a server you control.

Try these first

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

Upload [FAQ document], then ask the assistant: "What is your refund policy?" and compare the answer with the document.First try · your own server
Add our help center at [website address] as a web knowledge base.reads a website
Create a workflow app that searches the product manual first and answers only from it.
Embed the published assistant on [your website] with the embed code.public page

Example prompts written by Agent Flows Hub.

Features

From the creator's documentation.

Feature	What it does	Runs
Knowledge base	Upload PDF, DOCX, Markdown, spreadsheets and more, or add web pages; it splits them automatically	Your server
Model choice	Cloud models such as OpenAI, Claude and Gemini, or private models such as DeepSeek, Llama and Qwen	Model provider
Workflows	Visual workflows with knowledge search, conditions, tools and MCP	Your server
Publish and embed	Share a chat page or add the assistant to your website with embed code	The web
Images, audio and video	Text, image, audio and video input and output, with a model that supports them	Model provider
Paid editions	Chat app channels, single sign-on and the open API need a Professional or Enterprise license	Paid license

Good to know

You want customers or colleagues to get answers from your own manuals, policies or help pages. MaxKB splits uploaded documents into a knowledge base, connects it to the model you choose and publishes a chat app you can share or embed on a website.

Limits

The free community edition leaves out chat app channels such as WeCom, DingTalk, Lark and Slack, single sign-on and the open API, which need a paid Professional or Enterprise license. The maker's English pricing page still lists limits on users, apps and knowledge bases that its Chinese page says were removed, so check the current limits before you plan.
Answers depend on the model you connect and on how your documents are split. Ask questions you already know the answers to before you publish the app.

Notes by Agent Flows Hub (original project name: MaxKB), written from the creator's documentation, not from running it ourselves. MaxKB quick start · MaxKB README: Docker quick start · MaxKB editions compared (Chinese) · MaxKB pricing (English)

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, Linuxdocker run -d --name=maxkb --restart=always -p 8080:8080 -v ~/.maxkb:/opt/maxkb 1panel/maxkb
Windowsdocker run -d --name=maxkb --restart=always -p 8080:8080 -v C:/maxkb:/opt/maxkb registry.fit2cloud.com/maxkb/maxkb
Open itFollow the first-run steps in the app.
Give it a first taskUpload one short FAQ document, connect a model and ask a question whose answer is in it. Compare the reply with the document before sharing the app.

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.
Docker on a computer or server with at least 4 CPU cores, 8 GB of RAM and 100 GB of disk
An API key from a model provider such as OpenAI, or a local model through Ollama
A new admin password: change the default one after your first sign-in
Free. Nothing to pay.

Details

Made by: FIT2CLOUD
License: GPL-3.0 Terms
Platforms: macOS, Windows, Linux
Runs on: Self-hosted server

View source Official docs Report a problem

For AI agents

This page as a Markdown install brief:

/en/library/maxkb/install.md

Other apps

Maxun Turns a website into a table of data by recording a few clicks, and watches it for changes
Meetily Records your meetings, transcribes them live on your computer and writes a summary of each
MindsHub Hands research, reports and recurring tasks to AI agents in a web or desktop workspace
Ming Voice Runs a spoken voice assistant on your own graphics card, with timers, lists and smart home
MoneyPrinterTurbo Makes a short video with script, voiceover, subtitles and music from just a topic
Msty Studio Lets you chat with local and online AI models side by side in one private workspace

See all 113 apps

GitHub