RAG-Anything
New Active Docs only Last checked 2026-11-16 Permissive (MIT/Apache/BSD) CPU OK Developer library
๐ฌ One weekly email: local RAG releases, hardware compatibility updates, and practical setup guides. Free. Unsubscribe anytime.
Updates on X: @mkjerman
What it is
RAG-Anything from HKU's Data Intelligence Lab (the LightRAG team) is a multimodal RAG framework: it parses PDFs and Office files with MinerU (or Docling), splits out text, images, tables and equations, routes each through its own processor, and links them in a multimodal knowledge graph you can query with hybrid retrieval.
It's a developer library, not an app. README examples call OpenAI models; you supply your own LLM, vision-model and embedding functions. LightRAG itself now integrates RAG-Anything natively for multimodal ingest. RAGShelf has not run it.
- License
- MIT
- Latest release
- v1.4.4
- Release date
- 2026-10-28
- Hardware
- Python library (pip install raganything); built on LightRAG; MinerU parser needs Python 3.10โ3.13 (Docling parser alternative) unverified
- Document types
- PDF, Office (DOC/DOCX/PPT/PPTX/XLS/XLSX), images, text/Markdown unverified
- OCR / tables / retrieval
- Multimodal: dedicated processors for images, tables and equations; MinerU or Docling parsing; multimodal knowledge graph unverified
- Cloud calls
- Defaults to OpenAI in README examples (you pass the LLM, vision and embedding functions); local backends not covered in the README unverified
- Fully offline
- Needs config
- GitHub stars
- 24,565
- Last commit
- 2026-11-16 (last push, GitHub API 2026-11-16)
Install
Copied from the official README. Check the project docs for options.
pip install raganythingRetrieval
Graph RAG Hybrid (vector + keyword)
Based on: README: multimodal knowledge graph index + hybrid retrieval, built on LightRAG (checked 2026-11-09) (checked 2026-10-30). Vector vs. vectorless RAG explained โ
NVIDIA DGX Spark
DGX Spark: Unknown
No DGX Spark report found yet. We only mark a tool as working when there is a published source.
Last updated: 2026-11-16. License and release data from the GitHub API. Fields marked unverified have not yet been checked against official docs. How we know: Docs only (see the legend). Spot something wrong? Send a correction.
๐ฌ Get next week's tool
One new local RAG tool in your inbox every Monday. Free, unsubscribe anytime.
Sample issue: "AirLLM: run a 70B model on a 4GB GPU by loading one layer at a time. What it's good for (slow, private batch Q&A over your docs), what it isn't (fast chat), and the install command." Plus one new setup recipe or page, like our local vs. cloud cost calculator.