- Home
- Ai Databases
- Vald
Vald
What it is
Vald is a free, open-source, self-deployed engine for distributed approximate nearest-neighbor search over dense vectors. It uses the NGT algorithm by default, supports NGT and Faiss index types, and allows switching to another vector search engine. Asynchronous automatic indexing can continue while search requests are served. Its Kubernetes and cloud-native architecture supports horizontal scaling for memory and CPU. Vald supports index replication and can rebalance replicas automatically when an agent goes down. Backups can use Persistent Volumes, S3-compatible object storage, or both, with saved indexes restored when an agent pod starts. Ingress filters can vectorize or preprocess data, while egress filters post-process results; components must implement Vald’s gRPC interface. Official client libraries cover Go, Java, Python, and Node.js, with gRPC protobufs for other languages. Deployment documentation covers Kubernetes through Helm or its Helm operator, plus a Docker agent tutorial. Getting started requires a Kubernetes cluster, Helm v3 or later, and Go v1.15 or later. Kubernetes network policies can control cluster traffic when prerequisites are met.
For and against
For
- Free open-source project
- Asynchronous indexing serves searches
- Horizontal scaling for memory and CPU
- Replication and automatic rebalancing
- SDKs for Go, Java, Python, and Node.js
Against
- Requires a Kubernetes cluster
- Requires Helm v3 or later
- Getting started requires Go v1.15 or later
- Self-deployed software
The verdict
Vald provides distributed vector search with scaling, replication, backups, filters, and multiple SDKs. Its Kubernetes, Helm, and Go requirements make deployment a self-managed technical task.
Who it suits
Vald suits developers and infrastructure teams building vector search on Kubernetes or self-managed environments. It is relevant when SDKs, asynchronous indexing, replication, and configurable backups are needed.
Plans and prices
Vald
Open-source project · Self-deployed software
Read 10 Oct 2026 · vald.vdaas.org
On the record
- Backup
- Index backup can use Persistent Volumes, S3-compatible object storage, or both, and saved indexes can be restored when an agent pod starts.vald.vdaas.org · read 10 Oct 2026
- Deployment
- The maker documents deployment on Kubernetes with Helm or its Helm operator, and also provides a tutorial for running a Vald agent on Docker.vald.vdaas.org · read 10 Oct 2026
- Filtering
- Ingress filters can vectorize or preprocess data, while egress filters can post-process search results; filter components must implement Vald’s gRPC interface.vald.vdaas.org · read 10 Oct 2026
- Indexing
- Vald supports asynchronous automatic indexing while continuing to serve search requests during indexing.vald.vdaas.org · read 10 Oct 2026
- Product
- Vald is a highly scalable distributed approximate nearest neighbor engine for dense vector search.vald.vdaas.org · read 10 Oct 2026
- Requirements
- The getting-started guide requires a Kubernetes cluster, Helm v3 or later, and Go v1.15 or later.vald.vdaas.org · read 10 Oct 2026
- Resilience
- Vald supports index replication and automatically rebalances replicas when an agent goes down.vald.vdaas.org · read 10 Oct 2026
- Scaling
- Vald is Kubernetes and cloud-native based, with horizontal scaling for memory and CPU.vald.vdaas.org · read 10 Oct 2026
- SDKs and APIs
- Official client libraries are available for Go, Java, Python, and Node.js, and Vald publishes gRPC protobufs for other languages.vald.vdaas.org · read 10 Oct 2026
- Search algorithm
- Vald uses the NGT approximate nearest neighbor algorithm by default and says users can switch to another vector search engine.vald.vdaas.org · read 10 Oct 2026
- Security configuration
- Vald clusters can enable Kubernetes network policies to control ingress and egress traffic, subject to cluster prerequisites.vald.vdaas.org · read 10 Oct 2026
- Support
- The project recommends its community Slack for questions and also accepts email at [email protected].vald.vdaas.org · read 10 Oct 2026
