TusKANNy/seismic

Official repository of the Seismic library.

53
/ 100
Established

Seismic helps you quickly find relevant documents from large collections using sparse vector representations. It takes your existing document collection, represented as sparse vectors (e.g., keyword-score pairs), and allows you to query it with similar sparse vectors, returning the most similar documents and optionally their original text in microseconds. This is ideal for developers building search or recommendation systems that rely on learned sparse embeddings.

115 stars and 3 monthly downloads.

Use this if you need a high-performance, lightweight search engine for quickly retrieving documents based on learned sparse vector embeddings, especially for large datasets.

Not ideal if your data is not in sparse vector format or if you require exact similarity search results rather than fast approximate ones.

information-retrieval sparse-vector-search embedding-search RAG-pipelines recommender-systems
No Package No Dependents
Maintenance 13 / 25
Adoption 11 / 25
Maturity 16 / 25
Community 13 / 25

How are scores calculated?

Stars

115

Forks

12

Language

Rust

License

MIT

Last pushed

Mar 25, 2026

Monthly downloads

3

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/vector-db/TusKANNy/seismic"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.