[MLsys2026 Best Paper]: https://arxiv.org/abs/2506.08276. RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate, and 100% private RAG application on your personal device.
Python aifaissgpt-osslangchainllama-indexllmlocalstorageoffline-firstollamaprivacypythonragretrieval-augmented-generationvector-databasevector-searchvectors
add readme bench
yichuan5200309103201e3b571e7216
diff --git a/README.md b/README.mdindex 83ee8cd..14377fa 100755--- a/README.md+++ b/README.md@@ -372,23 +372,31 @@ python examples/run_evaluation.py data/indices/rpj_wiki/rpj_wiki.indexThe script will print the recall and search time for each query, followed by the average results.+### Storge Usage Comparison++| System | DPR(2M docs) | RPJ-wiki(60M docs) | Chat history() |+| --------------------- | ---------------- | ---------------- | ---------------- |+| Traditional Vector DB(LLamaindex faiss) | x GB | x GB | x GB |+| **Leann** | **xx MB** | **x GB** | **x GB** |+| **Reduction** | **x%** | **x%** | **x%** |+### Memory Usage Comparison-| System | 1M Documents | 10M Documents | 100M Documents |+| System | DPR(2M docs) | RPJ-wiki(60M docs) | Chat history() || --------------------- | ---------------- | ---------------- | ---------------- |-| Traditional Vector DB | 3.1 GB | 31 GB | 310 GB |-| **Leann** | **180 MB** | **1.2 GB** | **8.4 GB** |-| **Reduction** | **94.2%** | **96.1%** | **97.3%** |+| Traditional Vector DB(LLamaindex faiss) | x GB | x GB | x GB |+| **Leann** | **xx MB** | **x GB** | **x GB** |+| **Reduction** | **x%** | **x%** | **x%** |-### Query Performance+### Query Performance of LEANN-| Backend | Index Size | Query Time | Recall@10 |+| Backend | Index Size | Query Time | Recall@3 || ------------------- | ---------- | ---------- | --------- |-| DiskANN | 1M docs | 12ms | 0.95 |-| DiskANN + Recompute | 1M docs | 145ms | 0.98 |-| HNSW | 1M docs | 8ms | 0.93 |+| DiskANN | 1M docs | xms | 0.95 |+| HNSW | 1M docs | xms | 0.95 |++*Benchmarks run on Apple M3 Pro 36 GB*-*Benchmarks run on AMD Ryzen 7 with 32GB RAM*## 🏗️ Architecture