The smallest vector index in the world. LEANN to RAG Anything!

---

97% smaller than FAISS. RAG your emails, browser history, WeChat, or 60M documents on your laptop. No cloud, no API keys, no bullshit.

git clone https://github.com/yichuan520030910320/LEANN-RAG.git && cd LEANN-RAG
# 30 seconds later...
python demo.py  # RAG your first 1M documents

The Difference is Stunning

Bottom line: Index 60 million Wikipedia articles in 6GB instead of 201GB. Your MacBook can finally handle real datasets.

Why This Matters

Privacy: Your data never leaves your laptop. No OpenAI, no cloud, no "terms of service".

Speed: Real-time search on consumer hardware. No server setup, no configuration hell.

Scale: Handle datasets that would crash traditional vector DBs on your laptop.

30-Second Demo: RAG Your Life

from leann.api import LeannBuilder, LeannSearcher

# Index your entire email history (90K emails = 14MB vs 305MB)
builder = LeannBuilder(backend_name="hnsw")
builder.add_from_mailbox("~/Library/Mail")  # Your actual emails
builder.build_index("my_life.leann")

# Ask questions about your own data
searcher = LeannSearcher("my_life.leann") 
searcher.search("What did my boss say about the deadline?")
searcher.search("Find emails about vacation requests")
searcher.search("Show me all conversations with John about the project")

That's it. No cloud setup, no API keys, no "fine-tuning". Just your data, your questions, your laptop.

Try the interactive demo →

Get Started in 30 Seconds

Installation

git clone git@github.com:yichuan520030910320/LEANN-RAG.git leann
cd leann
git submodule update --init --recursive

macOS:

brew install llvm libomp boost protobuf
export CC=$(brew --prefix llvm)/bin/clang
export CXX=$(brew --prefix llvm)/bin/clang++
uv sync

Linux (Ubuntu/Debian):

sudo apt-get install libomp-dev libboost-all-dev protobuf-compiler libabsl-dev libmkl-full-dev libaio-dev
uv sync

Ollama Setup (Optional for Local LLM):

macOS:

First, download Ollama for macOS.

# Install Ollama
brew install ollama

# Pull a lightweight model (recommended for consumer hardware)
ollama pull llama3.2:1b

# For better performance but higher memory usage
ollama pull llama3.2:3b

Linux:

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh

# Start Ollama service manually
ollama serve &

# Pull a lightweight model (recommended for consumer hardware)
ollama pull llama3.2:1b

# For better performance but higher memory usage
ollama pull llama3.2:3b

Note: For Hugging Face models >1B parameters, you may encounter OOM errors on consumer hardware. Consider using smaller models like Qwen3-0.6B or switch to Ollama for better memory management.

Run the Demo (support .pdf,.txt,.docx, .pptx, .csv, .md etc)

uv run ./examples/main_cli_example.py

or you want to use python

source .venv/bin/activate
python ./examples/main_cli_example.py

Wild Things You Can Do

🕵️ Search Your Entire Life

python examples/mail_reader_leann.py
# "What did my boss say about the Christmas party last year?"
# "Find all emails from my mom about birthday plans"

90K emails → 14MB. Finally, search your email like you search Google.

🌐 Time Machine for the Web

python examples/google_history_reader_leann.py
# "What was that AI paper I read last month?"
# "Show me all the cooking videos I watched"

38K browser entries → 6MB. Your browser history becomes your personal search engine.

💬 WeChat Detective

python examples/wechat_history_reader_leann.py  
# "我想买魔术师约翰逊的球衣，给我一些对应聊天记录"
# "Show me all group chats about weekend plans"

400K messages → 64MB. Search years of chat history in any language.

📚 Personal Wikipedia

# Index 60M Wikipedia articles in 6GB (not 201GB)
python examples/build_massive_index.py --source wikipedia
# "Explain quantum computing like I'm 5"
# "What are the connections between philosophy and AI?"

PDF RAG Demo (using LlamaIndex for document parsing and Leann for indexing/search)

This demo showcases how to build a RAG system for PDF/md documents using Leann.

1. Place your PDF files (and other supported formats like .docx, .pptx, .xlsx) into the examples/data/ directory.
2. Ensure you have an OPENAI_API_KEY set in your environment variables or in a .env file for the LLM to function.

How It Works

LEANN doesn't store embeddings. Instead, it builds a lightweight graph and computes embeddings on-demand during search.

The magic: Most vector DBs store every single embedding (expensive). LEANN stores a pruned graph structure (cheap) and recomputes embeddings only when needed (fast).

Backends: DiskANN, HNSW, or FAISS - pick what works for your data size.

Performance: Real-time search on millions of documents. MLX support for 10-100x faster building on Apple Silicon.

Benchmarks

Run the comparison yourself:

python examples/compare_faiss_vs_leann.py

System	Storage
FAISS HNSW	5.5 MB
LEANN	0.5 MB
Savings	91%

Same dataset, same hardware, same embedding model. LEANN just works better.

Reproduce Our Results

uv pip install -e ".[dev]"  # Install dev dependencies
python examples/run_evaluation.py data/indices/dpr/dpr_diskann      # DPR dataset
python examples/run_evaluation.py data/indices/rpj_wiki/rpj_wiki.index  # Wikipedia

The evaluation script downloads data automatically on first run.

Storage Usage Comparison

System	DPR (2.1M chunks)	RPJ-wiki (60M chunks)	Chat history (400K messages)	Apple emails (90K messages chunks)	Google Search History (38K entries)
Traditional Vector DB(FAISS)	3.8 GB	201 GB	1.8G	305.8 MB	130.4 MB
LEANN	324 MB	6 GB	64 MB	14.8 MB	6.4MB
Reduction	91% smaller	97% smaller	97% smaller	95% smaller	95% smaller

Benchmarks run on Apple M3 Pro 36 GB

🏗️ Architecture

🔬 Paper

If you find Leann useful, please cite:

LEANN: A Low-Storage Vector Index

@misc{wang2025leannlowstoragevectorindex,
      title={LEANN: A Low-Storage Vector Index}, 
      author={Yichuan Wang and Shu Liu and Zhifei Li and Yongji Wu and Ziming Mao and Yilong Zhao and Xiao Yan and Zhiying Xu and Yang Zhou and Ion Stoica and Sewon Min and Matei Zaharia and Joseph E. Gonzalez},
      year={2025},
      eprint={2506.08276},
      archivePrefix={arXiv},
      primaryClass={cs.DB},
      url={https://arxiv.org/abs/2506.08276}, 
}

🤝 Contributing

We welcome contributions! Leann is built by the community, for the community.

Ways to Contribute

• 🐛 Bug Reports: Found an issue? Let us know!
• 💡 Feature Requests: Have an idea? We'd love to hear it!
• 🔧 Code Contributions: PRs welcome for all skill levels
• 📖 Documentation: Help make Leann more accessible
• 🧪 Benchmarks: Share your performance results

📈 Roadmap

🎯 Q2 2025

• DiskANN backend with MIPS/L2/Cosine support
• HNSW backend integration
• Real-time embedding pipeline
• Memory-efficient graph pruning

🚀 Q3 2025

• Advanced caching strategies
• Add contextual-retrieval https://www.anthropic.com/news/contextual-retrieval
• Add sleep-time-compute and summarize agent! to summarilze the file on computer!
• Add OpenAI recompute API

🌟 Q4 2025

• Integration with LangChain/LlamaIndex
• Visual similarity search
• Query rewrtiting, rerank and expansion

📄 License

MIT License - see LICENSE for details.

🙏 Acknowledgments

• Microsoft Research for the DiskANN algorithm
• Meta AI for FAISS and optimization insights
• HuggingFace for the transformer ecosystem
• Our amazing contributors who make this possible

---

⭐ Star us on GitHub if Leann is useful for your research or applications!

Made with ❤️ by the Leann team