Lance-bundle

Portable embeddings for local AI queries

AI AgentsRAGUnknownOpen source
Visit tool

github.com

About Lance-bundle

Lance-bundle is a library that packages precomputed embedding vectors alongside the actual embedding model for fast cold start to vector queries. It uses LanceDB + ONNX for low dependency footprint.

Description summarised by AI from the sources listed below.

Key features

  • Packages embedding vectors and the model that produced them into a single portable file
  • packaging precomputed embedding vectors
  • Allows for loading and querying it anywhere without a server or database
  • fast cold start to vector queries

Use cases

  • local AI queries
  • semantic queries to retrieve nearest documents/embeddings

Pricing

Pricing model: Unknown — we have not been able to confirm pricing from the official website, so nothing is stated here.

Pricing summarised by AI from the sources listed below.