EdgeVec
Sub-millisecond vector search in the browser (Rust/WASM)
Details
- External ID
- 46249896
- Source
- HN
- Company
- —
- Product
- EdgeVec
- Website domain
- github.com
- Launched
- Dec. 12, 2025
- Cohort
- —
- Upvotes
- 7
- Upvotes percentile
- 0.35877862595419846
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:25 p.m.
- Updated at
- Sept. 7, 2026, 9:25 p.m.
Description
Hi HN,I built EdgeVec, a vector database that runs entirely in the browser. It implements HNSW (Hierarchical Navigable Small World) graphs for approximate nearest neighbor search.Performance: - Sub-millisecond search at 100k vectors (768 dimensions, k=10) - 148 KB gzipped bundle - 3.6x memory reduction with scalar quantizationUse cases: browser extensions with semantic search, local-first apps, privacy-preserving RAG.Technical: Written in Rust, compiled to WASM. Uses AVX2 SIMD on native, simd128 on WASM. IndexedDB for browser persistence.npm: https://www.npmjs.com/package/edgevec GitHub: https://github.com/matte1782/edgevecThis is an alpha release. Main limitations: build time not optimized, no delete operations yet.Would love feedback from the community!
Enrichment
- Theme
- lightweight and on-device AI runtimes
- Vertical
- Horizontal
- Function
- Search & retrieval
- Audience
- Developer
- AI stance
- AI feature
- Project type
- Commercial product
- Normalized one-liner
- vector search in the browser
- Manually corrected
- False
Could you build this?
Partial Building a general web interface is easy, but implementing a production-grade HNSW graph vector search engine compiled from Rust to WASM with SIMD acceleration and tight memory constraints requires specialized algorithmic and systems engineering.
What it would actually take: The stack uses Rust compiled to WebAssembly with wasm-bindgen and WebAssembly SIMD primitives. The difficult part is implementing HNSW indexing and search algorithms with custom quantization (e.g., scalar or product quantization) to minimize browser memory overhead and guarantee sub-millisecond execution over 100k high-dimensional vectors.
Discussion
3 comments analyzed.
Competitors mentioned: Other in-browser vector libraries
Concerns raised: Search latency vs memory trade-offs comparison, How it compares to alternatives on performance metrics
Competitors
Other products that read as similar to this one — 120 launches clear the similarity bar, closest 8 shown.
Attention rank: #72 of 121 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 39 days after the earliest competitor.
- Zvec · hn · 2026-01-29 · 7 upvotes · similarity 0.50
- Semantic search over Hacker News, built on pgvector · hn · 2026-02-22 · 5 upvotes · similarity 0.47
- Optimizing LiteLLM with Rust · hn · 2025-11-18 · 27 upvotes · similarity 0.45
- SatoriDB · hn · 2025-12-30 · 5 upvotes · similarity 0.42
- real-time-retrieval · github · 2026-09-12 · 8 upvotes · similarity 0.42
- Minikv · hn · 2026-02-03 · 71 upvotes · similarity 0.42
- ParqDB · hn · 2026-08-21 · 26 upvotes · similarity 0.41
- Ipfrs · hn · 2026-01-20 · 5 upvotes · similarity 0.41
Other launches for this product
- No other launches for this product.