Vector IDB
A simple wrapper that turns IndexedDB into a vector database directly in the browser, for storing and querying vector data without a server.
Vector IDB explores using IndexedDB as a vector database inside the browser, providing a simple wrapper for storing and querying vector data with the VectorDB class. It was inspired by database companies specializing in vector search, such as Pinecone, and by other browser-based context tools.
A bundled demo loads real Wikipedia passages into IndexedDB, encodes a typed question with a local MiniLM encoder that needs no API key, and can search fully offline after the initial load; the project also publishes a research report on IndexedDB access patterns, HNSW indexing, WebGPU, and related architecture questions, generated from a Markdown source and checked with an automated citation and link check.
It is written in JavaScript and released under the Apache-2.0 license.