ragflow
RAGFlow is an open-source retrieval-augmented generation (RAG) engine that turns messy documents into a reliable context layer for LLMs. Its strength is deep document understanding: it parses complex PDFs, slides, spreadsheets, scanned images and web pages with layout analysis, OCR and table recognition before chunking, so answers can cite exactly where they came from.
It suits teams building internal knowledge bases, document Q&A and support assistants on top of their own data, especially when the source material is long, scanned or table-heavy. Template-based chunking can be inspected and adjusted by hand, answers come with traceable citations, and recent versions add agentic retrieval with selectable thinking depth plus knowledge compilation into wikis, graphs, mind maps and timelines. Data can be ingested from many sources, including websites via sitemaps and Google BigQuery.
Getting started: the recommended path is Docker Compose (Docker 24+, Compose 2.26.1+) on a machine with at least 4 CPU cores, 16 GB RAM and 50 GB of disk. After the services start you sign in through the web UI and configure an LLM, an embedding model and a reranker. A hosted cloud version is also available.
Features
Why choose it
Trade-offs
Maintenance check
- Last commit
- 2026-10-24
- Commits, last 90 days
- 100+
- Releases, last 90 days
- 6
- Latest release
- nightly (2025-12-01)
- Open issues
- 1340
- Recent issues with at least one comment
- 0/4
Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.
Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.
Most discussed open issues
- #4214 ROADMAP 202564 comments · 2024-12-24
- #11738 [Question]: After connecting to the Confluence data source, how do I pull Confluence data?35 comments · 2025-12-04
- #5598 [Question]: What type of GPU and how many GPUs are recommended to fasten the speed of parsing files?27 comments · 2025-03-04
- #7881 [Question]:when conducting a knowledge base search in the chat interface, it was found that ollama has a high cpu usage rate?26 comments · 2025-05-27
- #3022 [Question]: Hanging at parsing files24 comments · 2024-10-25
Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.
Who it fits
- Your documents have complex layouts (PDFs, scanned files, tables, slides) and you need layout analysis, OCR, and traceable citations.
- You want an Apache-2.0 all-in-one RAG platform where you can inspect and hand-tune template-based chunks in a web UI.
- Your server has less than 4 CPU cores and 16 GB RAM; RAGFlow's multi-service stack is heavy.
- You want an autonomous web research agent that writes 2,000-word reports from live search rather than indexing a local document corpus; GPT Researcher fits better.
How it differs
vs MaxKB★ 23.1kRAGFlow (Apache-2.0) specialises in deep document parsing (layout analysis, OCR, table extraction) and visual chunk editing. MaxKB (GPL-3.0) focuses on quick single-container Docker deployment, enterprise Q&A embedding, and domestic Chinese model integration.
vs GPT Researcher★ 30.3kRAGFlow is a document knowledge-base engine with layout-aware ingestion and cited Q&A; GPT Researcher is an autonomous multi-agent researcher that queries web search and local files to write long cited reports.
Compatibility
Quick start
Use cases
Alternatives
Featured in these lists
Comments
No comments yet. Be the first!