OpenAI

Chroma provides a convenient wrapper around OpenAI’s embedding API. This embedding function runs remotely on OpenAI’s servers, and requires an API key. You can get an API key by signing up for an account at OpenAI. The following OpenAI Embedding Models are supported:

text-embedding-ada-002
text-embedding-3-small
text-embedding-3-large

Visit OpenAI Embeddings documentation for more information.

Python
TypeScript

This embedding function relies on the openai python package, which you can install with pip install openai.You can pass in an optional model_name argument, which lets you choose which OpenAI embeddings model to use. By default, Chroma uses text-embedding-ada-002.

import chromadb.utils.embedding_functions as embedding_functions
openai_ef = embedding_functions.OpenAIEmbeddingFunction(
    api_key_env_var="OPENAI_API_KEY",
    model_name="text-embedding-3-small"
)

To use the OpenAI embedding models on other platforms such as Azure, you can use the api_base and api_type parameters:

import chromadb.utils.embedding_functions as embedding_functions
openai_ef = embedding_functions.OpenAIEmbeddingFunction(
    api_key_env_var="OPENAI_API_KEY",
    api_base="YOUR_API_BASE_PATH",
    api_type="azure",
    api_version="YOUR_API_VERSION",
    model_name="text-embedding-3-small"
)

You can pass in an optional model argument, which lets you choose which OpenAI embeddings model to use. By default, Chroma uses text-embedding-3-small.

// npm install @chroma-core/openai

import { OpenAIEmbeddingFunction } from "@chroma-core/openai";

const embeddingFunction = new OpenAIEmbeddingFunction({
    apiKeyEnvVar: "OPENAI_API_KEY",
    modelName: "text-embedding-3-small",
    // Optional: specify API base (e.g. for Azure OpenAI)
    apiBase: "your-api-base"
});

// use directly
const embeddings = embeddingFunction.generate(["document1", "document2"]);

// pass documents to query for .add and .query
let collection = await client.createCollection({
    name: "name",
    embeddingFunction: embeddingFunction,
});
collection = await client.getCollection({
    name: "name",
    embeddingFunction: embeddingFunction,
});

Dense Embedding Models

Sparse Embedding Models

Frameworks