Orama

// Overview

Orama is a search engine written in JavaScript. You define a schema, insert documents, and query them with full-text, vector, or hybrid search. The same code runs in the browser, on the edge, or on a server, so you can keep search close to where your data already lives.

It fits the vector-database category by storing embeddings directly in the index. You declare a vector field with a fixed size in the schema, insert documents with their embeddings, then search by passing a query vector and a similarity threshold. Full-text search uses BM25, and hybrid search combines both signals in one query.

It suits front-end and full-stack developers who want in-app or in-page search without standing up a separate search service. Extra features like filters, facets, geosearch, typo tolerance, and stemming in 30 languages cover common search needs out of the box.

// What it does

• Full-text search backed by BM25 ranking, with typo tolerance and exact match
• Vector and hybrid search by setting mode to 'vector' or 'hybrid' at query time
• Schema-typed documents with 10 field types, including vector[<size>], geopoint, and enum
• Runs in the browser, on the edge, on a server, or in Deno from the same package
• Filters, facets, field boosting, geosearch, and result pinning for merchandising
• Stemming and tokenization in 30 languages, plus a plugin system

// Getting started

Install the package, create a database with a schema, insert documents, then search.

npm i @orama/orama

import { create, insert, search } from '@orama/orama'

const db = create({
  schema: {
    name: 'string',
    description: 'string',
    price: 'number',
  },
})

insert(db, {
  name: 'Noise cancelling headphones',
  description: 'Best noise cancelling headphones on the market',
  price: 99.99,
})

const results = search(db, {
  term: 'Best headphones',
})

const results = search(db, {
  mode: 'vector',
  vector: {
    value: [0.938292, 0.284961, 0.248264, 0.748276, 0.26472],
    property: 'embedding',
  },
  similarity: 0.85,
})

Commands and code are distilled from the project's own documentation — always check the official repo for the latest.

// When to use it

• Add in-page search to a docs site or web app that runs entirely in the browser
• Build semantic or hybrid search over a product catalog using stored embeddings
• Run search on the edge or in serverless functions without a separate search service
• Power faceted, filtered product search with field boosting and result pinning

// How Orama compares

Orama alongside other open-source vector databases tools AI/TLDR tracks, ranked by GitHub stars.