Repolode
Submit
vespa-engine

vespa-engine/vespa

Vespa is an open-source search engine and serving platform for vectors, tensors, text and structured data at scale.

View on GitHubWebsite
Stars
7,141
Forks
747
Open issues
265
Last commit
6 hours ago

What is it?

Vespa is a platform for search, recommendation and personalization workloads. It selects a subset of a large, continuously changing data set, evaluates machine-learned models over the selected items, and organizes and returns the results, typically in under 100 milliseconds. The data is distributed over multiple nodes and evaluated in parallel.

The repository contains the full source code needed to build and run Vespa, and development happens here in the open. Applications can be deployed to Vespa Cloud or run on a self-managed Vespa instance. Custom logic can be added as Java components to a Vespa application.

The codebase is mostly Java, with C++ components built on AlmaLinux 8, and is licensed under Apache 2.0.

Key features

  • 01Serves vector, tensor, text and structured-data queries in one platform.
  • 02Evaluates machine-learned models over selected data at query time.
  • 03Distributes data and computation across multiple nodes for high availability.
  • 04Supports adding custom Java components to applications.
  • 05Can be run self-hosted or deployed to Vespa Cloud.

Use cases

  • Building search and retrieval systems that combine vector and keyword matching.
  • Running real-time recommendation and personalization services over large catalogs.
  • Serving retrieval-augmented generation (RAG) backends that need low-latency lookups.

Quick start

  1. 1

    Deploy your application to Vespa Cloud

  2. 2

    Or run your own Vespa instance by following the getting started guide

Things to know

  • Building the C++ and Java code is supported on AlmaLinux 8, and building the Java modules needs Java 17 and Maven 3.8+.
  • Running your own instance is more involved, and the README suggests adding more nodes for redundancy.

Topics

#search-engine#vector-search#recommendation#machine-learning#tensor#java#big-data#ai#rag#search

FAQ

What is vespa?
Vespa is an open-source search engine and serving platform for vectors, tensors, text and structured data at scale. Vespa is a platform for search, recommendation and personalization workloads. It selects a subset of a large, continuously changing data set, evaluates machine-learned models over the selected items, and organizes and returns the results, typically in under 100 milliseconds. The data is distributed over multiple nodes and evaluated in parallel.
Is vespa free to use? What license does it have?
vespa is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is vespa still maintained?
Actively maintained. The last commit to vespa was on Oct 22, 2026.
What can I use vespa for?
Building search and retrieval systems that combine vector and keyword matching. Running real-time recommendation and personalization services over large catalogs. Serving retrieval-augmented generation (RAG) backends that need low-latency lookups.
How do I install vespa?
Deploy your application to Vespa Cloud Or run your own Vespa instance by following the getting started guide
Where can I get vespa?
The source code and installation instructions are on GitHub: https://github.com/vespa-engine/vespa

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects