lightonai/fast-plaid
High-Performance Engine for Multi-Vector Search
Implements per-token late interaction search via K-means-clustered centroid indexing, enabling fine-grained token-level similarity matching instead of single fixed-size embeddings. Written in Rust with GPU acceleration and exposed through PyO₃ bindings for seamless PyTorch integration. Supports dynamic index updates with automatic centroid expansion, metadata filtering via document subsets, and token-level similarity matrices for models like ColBERT and ColPali.
218 stars and 61,035 monthly downloads. Used by 1 other package. Available on PyPI.
Stars
218
Forks
17
Language
Python
License
MIT
Category
Last pushed
Mar 06, 2026
Monthly downloads
61,035
Commits (30d)
0
Dependencies
8
Reverse dependents
1
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/vector-db/lightonai/fast-plaid"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
databendlabs/databend
Data Agent Ready Warehouse : One for Analytics, Search, AI, Python Sandbox. — rebuilt from...
oceanbase/oceanbase
The Fastest Distributed Database for Transactional, Analytical, and AI Workloads.
matrixorigin/matrixone
MySQL-compatible HTAP database with Git for Data, vector search, and fulltext search....
ArcadeData/arcadedb
ArcadeDB Multi-Model Database, one DBMS that supports SQL, Cypher, Gremlin, HTTP/JSON, MongoDB...
HelixDB/helix-db
HelixDB is an open-source graph-vector database built from scratch in Rust.