Orama

Orama is a lightweight, open-source search engine supporting full-text, vector, and hybrid search. It is suitable for use in browsers, servers, or edge environments, with a very small footprint (less than 2kb).

Features

• Lightweight: Core library is under 2kb, making it suitable for resource-constrained environments.
• Multi-environment support: Can be used in browsers, on servers, or at the edge.
• Search Modes:
  • Full-text search
  • Vector search
  • Hybrid search (combines full-text and vector search)
• RAG Pipeline: Supports Retrieval-Augmented Generation (RAG) workflows for building chat or search experiences similar to ChatGPT or Perplexity.
• Flexible Data Types: Supports 10 data types in its indexing schema:
  • string, number, boolean, enum, geopoint, string[], number[], boolean[], enum[], vector[size]
• Embeddings Support: Ability to use text embeddings for vector and hybrid search. Plugins available for generating embeddings and proxying OpenAI embedding models securely from the client side.
• Plugin System: Official plugins available and support for writing custom plugins.
• Open Source: Licensed under the Apache 2.0 license.

Tags

open-source, vector-search, hybrid-search, lightweight

Category

Vector Database Engines

Pricing

Orama is open-source software and free to use under the Apache 2.0 license.