Orama
Full-text, vector, and hybrid search with a unique API.
On your browser, server, mobile app, or at the edge.
In less than 2kb.
Join Orama's Slack channel
If you need more info, help, or want to provide general feedback on Orama, join
the
Orama Slack channel
Highlighted features
Installation
You can install Orama using npm
, yarn
, pnpm
, bun
:
npm i @orama/orama
Or import it directly in a browser module:
<html>
<body>
<script type="module">
import { create, search, insert } from 'https://unpkg.com/@orama/orama@latest/dist/index.js'
</script>
</body>
</html>
With Deno, you can just use the same CDN URL or use npm specifiers:
import { create, search, insert } from 'npm:@orama/orama'
Read the complete documentation at https://docs.askorama.ai.
Usage
Orama is quite simple to use. The first thing to do is to create a new database
instance and set an indexing schema:
import { create, insert, remove, search, searchVector } from '@orama/orama'
const db = await create({
schema: {
name: 'string',
description: 'string',
price: 'number',
embedding: 'vector[1536]',
meta: {
rating: 'number',
},
},
})
Orama currently supports 10 different data types:
Type | Description | example |
---|
string | A string of characters. | 'Hello world' |
number | A numeric value, either float or integer. | 42 |
boolean | A boolean value. | true |
enum | An enum value. | 'drama' |
geopoint | A geopoint value. | { lat: 40.7128, lon: 74.0060 } |
string[] | An array of strings. | ['red', 'green', 'blue'] |
number[] | An array of numbers. | [42, 91, 28.5] |
boolean[] | An array of booleans. | [true, false, false] |
enum[] | An array of enums. | ['comedy', 'action', 'romance'] |
vector[<size>] | A vector of numbers to perform vector search on. | [0.403, 0.192, 0.830] |
Orama will only index properties specified in the schema but will allow you to set and store additional data if needed.
Once the db instance is created, you can start adding some documents:
await insert(db, {
name: 'Wireless Headphones',
description: 'Experience immersive sound quality with these noise-cancelling wireless headphones.',
price: 99.99,
embedding: [...],
meta: {
rating: 4.5,
},
})
await insert(db, {
name: 'Smart LED Bulb',
description: 'Control the lighting in your home with this energy-efficient smart LED bulb, compatible with most smart home systems.',
price: 24.99,
embedding: [...],
meta: {
rating: 4.3,
},
})
await insert(db, {
name: 'Portable Charger',
description: 'Never run out of power on-the-go with this compact and fast-charging portable charger for your devices.',
price: 29.99,
embedding: [...],
meta: {
rating: 3.6,
},
})
After the data has been inserted, you can finally start to query the database.
const searchResult = await search(db, {
term: 'headphones',
})
In the case above, you will be searching for all the documents containing the
word "headphones"
, looking up in every string
property specified in the schema:
{
elapsed: {
raw: 99512,
formatted: '99μs',
},
hits: [
{
id: '41013877-56',
score: 0.925085832971998432,
document: {
name: 'Wireless Headphones',
description: 'Experience immersive sound quality with these noise-cancelling wireless headphones.',
price: 99.99,
meta: {
rating: 4.5
}
}
}
],
count: 1
}
You can also restrict the lookup to a specific property:
const searchResult = await search(db, {
term: 'immersive sound quality',
properties: ['description'],
})
Result:
{
elapsed: {
raw: 21492,
formatted: '21μs',
},
hits: [
{
id: '41013877-56',
score: 0.925085832971998432,
document: {
name: 'Wireless Headphones',
description: 'Experience immersive sound quality with these noise-cancelling wireless headphones.',
price: 99.99,
meta: {
rating: 4.5
}
}
}
],
count: 1
}
You can use non-string data to filter, group, and create facets:
const searchResult = await search(db, {
term: 'immersive sound quality',
where: {
price: {
lte: 199.99
},
rating: {
gt: 4
}
},
})
Performing hybrid and vector search
Orama is a full-text and vector search engine. This allows you to adopt different kinds of search paradigms depending on your specific use case.
To perform vector or hybrid search, you can use the same search
method used for full-text search.
You'll just have to specify which property you want to perform vector search on, and a vector to be used to perform vector similarity:
const searchResult = await searchVector(db, {
mode: 'vector',
vector: {
value: [...],
property: 'embedding'
}
})
If you're using the Orama Secure AI Proxy (highly recommended), you can skip the vector configuration at search time, since the official Orama Secure AI Proxy plugin will take care of it automatically for you:
import { create } from '@orama/orama'
import { pluginSecureProxy } from '@orama/plugin-secure-proxy'
const secureProxy = secureProxyPlugin({
apiKey: '<YOUR-PUBLIC-API-KEY>',
defaultProperty: 'embedding',
model: 'openai/text-embedding-ada-002'
})
const db = await create({
schema: {
name: 'string',
description: 'string',
price: 'number',
embedding: 'vector[1536]',
meta: {
rating: 'number',
},
},
plugins: [secureProxy]
})
const resultsHybrid = await search(db, {
mode: 'vector',
term: 'Videogame for little kids with a passion about ice cream',
where: {
price: {
lte: 19.99
},
'meta.rating': {
gte: 4.5
}
}
})
Performing Geosearch
Orama supports Geosearch as a search filter. It will search through all the properties specified as geopoint
in the schema:
import { create, insert } from '@orama/orama'
const db = await create({
schema: {
name: 'string',
location: 'geopoint'
}
})
await insert(db, { name: 'Duomo di Milano', location: { lat: 45.46409, lon: 9.19192 } })
await insert(db, { name: 'Piazza Duomo', location: { lat: 45.46416, lon: 9.18945 } })
await insert(db, { name: 'Piazzetta Reale', location: { lat: 45.46339, lon: 9.19092 } })
const searchResult = await search(db, {
term: 'Duomo',
where: {
location: {
radius: {
coordinates: {
lat: 45.4648,
lon: 9.18998
},
unit: 'm',
value: 1000,
inside: true
}
}
}
})
Orama Geosearch APIs support distance-based search (via radius
), or polygon-based search (via polygon
).
By default, Orama will use the Haversine formula to perform Geosearch, but high-precision search can be enabled by passing the highPrecision
option in your radius
or polygon
configuration. This will tell Orama to use the Vicenty Formulae instead, which is more precise for longer distances.
Read more in the official docs.
Official Docs
Read the complete documentation at https://docs.askorama.ai.
Official Orama Plugins
Write your own plugin: https://docs.askorama.ai/open-source/plugins/writing-your-own-plugins
License
Orama is licensed under the Apache 2.0 license.