Prerequisites
Before installing VecLabs, ensure you have:- Node.js 16+ (for TypeScript) or Python 3.10+ (for Python)
- No Solana wallet required for basic operations (upsert, query, delete)
- Solana wallet needed for on-chain verification features (optional)
Vectors are currently stored in-memory. Persistent decentralized storage via Shadow Drive is in active development and ships in v0.2.0.
Install the SDK
VecLabs provides SDKs for both TypeScript and Python with identical APIs.The TypeScript package is published as
solvec on npm (currently alpha: solvec@alpha). The Python package is solvec on PyPI with the --pre flag for alpha releases.Verify Installation
Verify your installation by checking the package version:Network Configuration
VecLabs supports three Solana networks:Mainnet Beta
Production environment for live applications
Devnet
Development environment for testing (recommended for getting started)
Localnet
Local Solana validator for offline development
Optional: Solana Wallet Setup
For on-chain verification features, you’ll need a Solana wallet:Install Solana CLI
Follow the official Solana installation guide or use:
Custom RPC Endpoint
For production applications, consider using a dedicated RPC provider:Current Status
VecLabs is in alpha with a stable API surface:| Feature | Status |
|---|---|
| upsert / query / delete / fetch | ✅ Working |
| Cosine, euclidean, dot product metrics | ✅ Working |
| Metadata filtering | ✅ Working |
| Merkle root computation | ✅ Working |
| verify() method | ✅ Working (local computation) |
| Solana on-chain Merkle updates | 🚧 In progress |
| Shadow Drive persistence | 🚧 In progress (in-memory for now) |
| WASM Rust HNSW bridge | 🚧 In progress (JS fallback for now) |
Links & Resources
GitHub Repository
View source code and contribute
Solana Program
Live on Solana devnet
npm Package
TypeScript SDK on npm
PyPI Package
Python SDK on PyPI
Next Steps
Quickstart Guide
Learn how to create your first vector collection and perform searches