docs

BrowserVec developer docs

Start here, then follow the path that matches what you're doing.

I'm new to BrowserVec

• Beginner guide — what's a vector? what's similarity search? plain-language walkthrough, no ANN jargon
• Quick start — get the demo running in 2 commands
• Features — overview of every capability
• Core API — minimal code example to see the shape
• Benchmarks — what performance to expect

I'm integrating BrowserVec into my app

• Integration guide — install, bundle, CSP, browser support, error handling, device reuse, lifecycle
• Configuration guide — decision tree, parameter tuning, per-query overrides
• API reference — every method, config field, and type
• Feature deep-dives — code examples for each feature

I'm choosing an index configuration

• Decision tree → — flat fp32, quantized, IVF, or IVF×quant combo
• Parameter tuning — dimension, metric, quantBits, rerankFactor, nprobe, nlist

I want to understand how it works

• Architecture — module map, data flow, index selection, ingest/query/delete paths, file↔spec mapping table
• Internals — device acquisition, store internals, persistence format, encryption envelope, quantization codec, IVF/k-means details, Worker offload seam, CPU/WASM fallback details

I want to run the examples

Open any via npm run dev:

Example	Feature
examples/01-basic-flat.html	Basic flat search — create, addBatch, query
examples/02-quantization.html	int8 / int4 quantization comparison
examples/03-ivf-index.html	IVF approximate search with nprobe sweep
examples/04-ivf-quant-combo.html	IVF × int8/int4 combo — the 1M path
examples/18-hnsw-index.html	HNSW graph index — efSearch sweep, recall vs speed, works without WebGPU
examples/19-hnsw-gpu.html	GPU graph search — one-dispatch beam kernel, queryBatch, CPU vs GPU crossover
examples/05-text-retrieval.html	End-to-end text → search
examples/06-persistence.html	Save/load to OPFS/IndexedDB
examples/07-encryption.html	AES-256-GCM encrypted snapshots
examples/08-deletion-compaction.html	Tombstone deletes, update, compact
examples/09-worker-ingest.html	Worker-offloaded quantized ingest
examples/10-chunking.html	Multi-buffer corpus chunking
examples/11-custom-embedder.html	Custom embedder implementation
examples/12-transformers-embedder.html	Transformers.js semantic embeddings
examples/13-all-metrics.html	Cosine vs dot vs L2 comparison
examples/14-advanced-api.html	isSupported, stats, destroy, device reuse
examples/15-browser-extension.html	Browser extension pattern — capture pages, persist, semantic search, offline
examples/16-react-hooks.html	React hooks: useVectorStore, useSimilaritySearch, useEmbedding, useRetriever
examples/17-vue-composables.html	Vue composables: useVectorStore, useSimilaritySearch, useEmbedding, useRetriever
examples/perf-benchmark.html	Large-scale performance benchmark

I want to contribute to BrowserVec

• Development guide — setup, build, project layout, how to add an index type / WGSL kernel, code conventions
• Architecture — understand the module structure and data flow before making changes

Reference

• API reference — complete public API surface
• Architecture → spec mapping — which file implements which REQUIREMENTS.md section
• REQUIREMENTS.md — original design spec
• CHANGELOG.md — release history