// rag · pingcap/autoflow

autoflow

autoflow is an open-source rag & knowledge project by pingcap (pingcap/autoflow on GitHub). It has 2,978 GitHub stars and 194 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 27 April 2026.

Graph RAG conversational knowledge base built on TiDB Serverless vector storage

★ 3.0k stars⑂ 194 forksApache-2.0updated 2026-04

View on GitHub ↗Official site ↗

▦ visual — this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.

// project facts

Project facts

Repository: pingcap/autoflow
GitHub stars: 2,978
Forks: 194
License: Apache-2.0
Last updated on GitHub: 27 April 2026
Category: RAG & Knowledge
Listing: Full install guide
Official site: Official site

// how it works

How it works

Clarify your goal
Decide what knowledge you want to make conversational — typically documentation or a website that the built-in crawler can ingest via sitemap scraping.
Get AutoFlow
Clone the repo and read the deployment docs linked from the README.
Install and configure
Deploy with Docker Compose (the docs suggest 4 CPU cores and 8 GB RAM) backed by TiDB Serverless vector storage, and configure your LLM settings.
Run your first task
Open the conversational search page, ask a question about your ingested content, and optionally embed the chat snippet on your site.

// features

Why people love it

Graph RAG knowledge base

Retrieval is built on a knowledge-graph RAG approach using TiDB Vector storage, via LlamaIndex and DSPy.

Perplexity-style search page

A conversational search interface for asking questions over your knowledge base.

Built-in website crawler

Crawls official and documentation sites through sitemap URL scraping to build coverage.

Embeddable JS snippet

Drop a simple JavaScript snippet into your website to add a conversational search window, typically at the bottom right.

TiDB-backed storage

TiDB stores chat history, vectors, JSON and analytics in one database.

Docker Compose deployment

Documented deployment path with Docker Compose in the project's deployment docs.

// run it

Run it yourself

Clone the repo

Clone AutoFlow and enter the directory.

git clone https://github.com/pingcap/autoflow.git
cd autoflow

Deploy with Docker Compose
Follow the deployment docs linked in the README for the Docker Compose deployment (4 CPU / 8 GB RAM suggested).
```
# see README for the exact command — deployment docs: https://autoflow.tidb.ai/deploy-with-docker
```
Try the live demo first
If you want to see it before deploying, the hosted demo runs at tidb.ai.
```
# open https://tidb.ai
```

// use cases

What you can build

Conversational search over product documentation
A support knowledge base that answers from your own site content
Embedding a Q&A widget on a documentation or marketing site
Experimenting with Graph RAG on TiDB Serverless

For a different angle on this shelf, dstoolkit-text2sql-and-imageprocessing opens rag & knowledge for this repo accelerates development of rag applications w. Also on this shelf: tela — a self-hostable markdown team wiki with a built-in MCP server.

// faq

Questions, answered

Who maintains AutoFlow?

PingCAP, the company behind TiDB, maintains AutoFlow as an open-source project.

Is AutoFlow free?

The code is Apache-2.0 licensed and free to use. You will typically use TiDB Serverless and an LLM provider, whose own pricing applies.

Does it need a paid API key?

It is an LLM-based RAG tool, so you should expect to configure an LLM provider; TiDB Serverless is used for vector storage. Check the deployment docs for the exact configuration.

How does it compare to plain vector RAG?

AutoFlow adds a knowledge-graph (Graph RAG) layer and a ready conversational UI with crawling and an embeddable widget, rather than just a retrieval library.

How do I get started?

The most-starred other RAG & Knowledge projects in the LibreAI catalogue — the realistic shortlist if autoflow is not the right fit.

MarkItDown★ 193.5k
MarkItDown — Microsoft tool that converts PDFs, Office files, images and more to clean Markdown for LLMs, MIT
Dify★ 157.8k
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or…
ragflow★ 91.6k
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context…
PaddleOCR★ 90.8k
Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages
MinerU★ 82.5k
MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms

// more in rag

Get autoflow

Free and open — grab the source, star the repo, and start creating.

View on GitHub ↗← More RAG & Knowledge tools

autoflow

Project facts

How it works

Why people love it

Graph RAG knowledge base

Perplexity-style search page

Built-in website crawler

Embeddable JS snippet

TiDB-backed storage

Docker Compose deployment

Run it yourself

What you can build

Questions, answered

Similar tools

helix-db

Langchain-Chatchat

algebrica

SAG

code-graph-rag

qdrant-loader

Alternatives in RAG & Knowledge

More in RAG & Knowledge

Get autoflow