# Vald

> Vald is a distributed vector search engine built for high scalability and low latency, using NGram-based filtering and Go implementation. It supports sharding and high availability for cloud-native deployments. Suited for real-time recommendations; similar to Milvus but lighter with focus on NG-Tree indexing vs full feature set.

- **URL:** https://vald.vdaum.net/
- **Category:** Scalable Distributed Vector DBs
- **Tags:** Distributed Search, Ngt Index, Go Lang
- **Updated:** 2026-04-29 11:31
- **Canonical page:** https://vectordb.ever.works/items/vald

## Details

## Overview

Scalable vector search middleware.

## Features

- Auto scaling
- Fault tolerant
- gRPC and REST APIs

## Pricing

Free and open-source.

---

_This Markdown mirror is generated by Ever Works for AI agents. The canonical HTML page is at https://vectordb.ever.works/items/vald._