{"version":1,"type":"story","url":"https://digestai.news/story/meet-leann-an-open-source-vector-index-that-shrinks-a-60m-chunk-rag-in","json":"https://digestai.news/story/meet-leann-an-open-source-vector-index-that-shrinks-a-60m-chunk-rag-in.json","markdown":"https://digestai.news/story/meet-leann-an-open-source-vector-index-that-shrinks-a-60m-chunk-rag-in.md","slug":"meet-leann-an-open-source-vector-index-that-shrinks-a-60m-chunk-rag-in","headline":"Meet LEANN: An Open-Source Vector Index that Shrinks a 60M-Chunk RAG Index from 201 GB to 6 GB","summary":"UC Berkeley’s Sky Computing Lab introduced LEANN, an open-source vector index that reduces storage needs by recomputing embeddings at query time instead of storing them. A 60 million-chunk Wikipedia index shrinks from 201 GB to 6 GB in the repo benchmark, and a 76 GB corpus setup drops from 173 GB (embeddings) + 15 GB (graph) to 4 GB in the paper’s tests. The project won the Best Research Paper award at MLSys 2026.\n\nLEANN achieves this by storing only a pruned graph and a compact product quantization table. At query time, it re-embeds only nodes on the search path, using techniques like two-level search, dynamic batching, and high-degree preserving pruning. Performance matches HNSW’s downstream accuracy at 90% recall while using under 5% of raw-data storage. Benchmarks show LEANN’s retrieval time ranges from 1.12 seconds (GPQA) to 7.12 seconds (HotpotQA) on an RTX 4090, compared to HNSW’s 0.05 seconds. The MIT-licensed tool runs on laptops and workstations, with apps for documents, Apple Mail, Chrome history, and more.","keyPoints":["LEANN reduces a 60M-chunk Wikipedia index from 201 GB to 6 GB by recomputing embeddings on demand","Matches HNSW accuracy at 90% recall while using 4 GB instead of 188 GB for a 60M-passage datastore","MIT-licensed open-source tool with apps for documents, Apple Mail, and Slack/Twitter data retrieval"],"whyItMatters":"LEANN enables semantic search and RAG on personal data without bloating storage, making it practical for laptops and workstations.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["UC Berkeley","Sky Computing Lab","MarkTechPost","Marktechpost AI Media Inc."],"models":[],"people":["Asif Razzaq"]},"firstPublishedAt":"2026-10-27T21:54:37Z","updatedAt":"2026-10-27T21:54:37Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"MarkTechPost","title":"Meet LEANN: An Open-Source Vector Index that Shrinks a 60M-Chunk RAG Index from 201 GB to 6 GB","url":"https://marktechpost.com/2026/10/27/leann-vector-index-mlsys-2026","publishedAt":"2026-10-27T21:54:37Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Meet LEANN: An Open-Source Vector Index that Shrinks a 60M-Chunk RAG Index from 201 GB to 6 GB\", 27 October 2026, https://digestai.news/story/meet-leann-an-open-source-vector-index-that-shrinks-a-60m-chunk-rag-in","publisher":"Digest AI","title":"Meet LEANN: An Open-Source Vector Index that Shrinks a 60M-Chunk RAG Index from 201 GB to 6 GB","datePublished":"2026-10-27T21:54:37Z","url":"https://digestai.news/story/meet-leann-an-open-source-vector-index-that-shrinks-a-60m-chunk-rag-in"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}