Vald

Vector Databases

Vald is an open-source, cloud-native vector search engine designed for approximate nearest neighbor (ANN) searches on dense vector data.

Reviewed by 7wData
Share Email

On this page

Profile

Vald is a distributed, cloud-native vector database that performs fast approximate nearest neighbor searches on large-scale vector data.

Vald is an open-source, cloud-native vector search engine designed for approximate nearest neighbor (ANN) searches on dense vector data. It was created by VDAAS (Vector Database as a Service) and is built on the NGT (Neighborhood Graph and Tree) algorithm originally developed by Yahoo Japan. Vald is written in Go and runs on Kubernetes, offering features like asynchronous auto-indexing, horizontal scaling, index replication, and automated backup to object storage.

The project lists LY Corporation (formerly Yahoo Japan) and the Japan Search portal (jpsearch.go.jp) as users. Vald provides official client SDKs for Go, Java, Python, Node.js, and Clojure. The project is distributed under an open-source license and does not appear to have disclosed any venture capital funding, revenue, or headcount figures.

Its primary distribution channel is via GitHub and Helm charts for Kubernetes deployment. The most recent publicly available information is the project's website, which has not been updated with any news or financial data since its initial release. No funding rounds, revenue results, or major product announcements have been reported by independent outlets like TechCrunch or The Information within the last year. The project's GitHub activity and community engagement are not detailed in the provided research dossier.

Track Vald and 240+ vendors.

48k+ subscribers read the daily AI & data note. One email, both newsletters. Unsubscribe anytime.

Who buys this

  • Organizations building AI-powered search and recommendation systems
  • Companies needing scalable similarity search for image, video, or text embeddings
  • Enterprises deploying machine learning pipelines that require vector indexing
  • Developers seeking an open-source alternative to proprietary vector databases

Publicly disclosed clients

  • LY Corporation
  • Japan Search (jpsearch.go.jp)

Strengths and what to watch

Strengths

  • Cloud-native architecture designed for Kubernetes, enabling horizontal scaling and automated operations.
  • Uses the NGT algorithm, a high-performance ANN library originally developed by Yahoo Japan, for fast and accurate vector search.
  • Open-source with a permissive license and multi-language SDK support, reducing vendor lock-in.

Watch for

  • No disclosed funding, revenue, or headcount; the project's long-term sustainability and development velocity are unclear.
  • Very limited publicly known customer base (only two named users), raising questions about real-world adoption and production maturity.
  • No independent news coverage or financial results in the past year; the project may have low community or commercial traction.

Key Information

Industry
Vector Databases
Founded
2015
Headquarters
Brisbane, Australia

Frequently Asked Questions

What is Vald and what does it do?

Vald is an open-source, cloud-native vector database designed for fast approximate nearest neighbor searches on large-scale vector data. It runs on Kubernetes and uses the NGT algorithm originally developed by Yahoo Japan to perform similarity searches on dense vectors.

How does Vald handle vector search?

Vald performs approximate nearest neighbor searches using the NGT algorithm, a high-performance library originally built by Yahoo Japan. This allows it to quickly find similar vectors in large datasets, making it suitable for AI-powered search and recommendation systems.

Can Vald be deployed on Kubernetes?

Yes, Vald is built specifically for Kubernetes, offering a cloud-native architecture. It supports horizontal scaling, asynchronous auto-indexing, index replication, and automated backup to object storage, making it easy to manage in containerized environments.

What programming languages does Vald support?

Vald provides official client SDKs for Go, Java, Python, Node.js, and Clojure. This multi-language support allows developers to integrate vector search into applications built with various programming languages without vendor lock-in.

Who are the known users of Vald?

Vald lists LY Corporation, formerly Yahoo Japan, and the Japan Search portal as its known users. However, the project has only these two publicly named customers, which raises questions about its broader adoption and production maturity in the industry.

Is Vald a good open-source alternative to proprietary vector databases?

Vald is open-source with a permissive license and offers cloud-native features like Kubernetes deployment and multi-language SDKs. However, its limited known user base and lack of disclosed funding or revenue may affect its long-term sustainability compared to more established proprietary options.

Sources

  1. vald.vdaas.org — Product description, features, base technologies, SDKs, and listed users (LY Corporation, Japan Search).
  2. investorvalero.com — This URL is for Valero Energy (NYSE: VLO), not Vald. It was incorrectly included in the research dossier and does not support any facts about Vald.
  3. www.elevancehealth.com — This URL is for Elevance Health (NYSE: ELV), not Vald. It was incorrectly included in the research dossier and does not support any facts about Vald.
  4. www.novonordisk.com — This URL is for Novo Nordisk, not Vald. It was incorrectly included in the research dossier and does not support any facts about Vald.
  5. techcrunch.com — TechCrunch homepage accessed on 2026-06-05; no stories about Vald were found.
  6. techcrunch.com — TechCrunch article on Q1 2025 startup funding trends; does not mention Vald.
  7. techcrunch.com — TechCrunch funding tag page; no stories about Vald were found.