Quickstart

Get up and running with Chroma in minutes. This guide will walk you through installing Chroma and creating your first vector database.

Installation

pip install chromadb

Basic example

Python
JavaScript

import chromadb

# Create a Chroma client
client = chromadb.Client()

# Create a collection
collection = client.create_collection(name="my_collection")

# Add documents
collection.add(
    documents=[
        "This is a document about pineapple",
        "This is a document about oranges"
    ],
    ids=["id1", "id2"]
)

# Query the collection
results = collection.query(
    query_texts=["This is a query about fruit"],
    n_results=2
)

print(results)

import { ChromaClient } from "chromadb";

// Create a Chroma client
const client = new ChromaClient();

// Create a collection
const collection = await client.createCollection({
  name: "my_collection"
});

// Add documents
await collection.add({
  ids: ["id1", "id2"],
  documents: [
    "This is a document about pineapple",
    "This is a document about oranges"
  ]
});

// Query the collection
const results = await collection.query({
  queryTexts: ["This is a query about fruit"],
  nResults: 2
});

console.log(results);

Expected output

{
  'ids': [['id1', 'id2']],
  'distances': [[1.0, 1.2]],
  'documents': [['This is a document about pineapple', 
                 'This is a document about oranges']],
  'metadatas': [[None, None]]
}

Chroma automatically:

Generates embeddings using the default model
Stores embeddings in the collection
Finds the most similar documents when you query

What’s happening?

Create a client

The client connects to Chroma. By default, it runs in-memory for quick prototyping.

Create a collection

Collections store your embeddings, documents, and metadata.

Add documents

Chroma automatically generates embeddings for your documents using the default embedding function.

Query

Query with natural language. Chroma finds the most similar documents using vector similarity search.

Client types

Choose the right client for your use case:

In-Memory (Default)
Persistent
Client-Server
Chroma Cloud

import chromadb

# Ephemeral client - data is cleared when program exits
client = chromadb.Client()

Perfect for: Testing, prototyping, Jupyter notebooks

import chromadb

# Data persists to disk
client = chromadb.PersistentClient(path="./chroma_db")

Perfect for: Local development, small applications

import chromadb

# Connect to a Chroma server
client = chromadb.HttpClient(host="localhost", port=8000)

Perfect for: Production deployments, distributed systems

Run the server with: chroma run --path ./chroma_db

import chromadb

# Connect to Chroma Cloud
client = chromadb.CloudClient(
    tenant="my-tenant",
    database="my-database",
    api_key="your-api-key"
)

Perfect for: Serverless, managed deployments

Next steps

Core Concepts

Learn about collections, embeddings, and metadata

Embedding Functions

Use OpenAI, Cohere, or custom embeddings

Filtering

Filter results using metadata

Deployment

Deploy Chroma to production

Get Started

Core Concepts

Guides

Deployment

Operations

Installation

Basic example

Expected output

What’s happening?

Client types

Next steps

Core Concepts

Embedding Functions

Filtering

Deployment

Build docs developers (and LLMs) love

Get Started

Core Concepts

Guides

Deployment

Operations

​Installation

​Basic example

​Expected output

​What’s happening?

​Client types

​Next steps

Core Concepts

Embedding Functions

Filtering

Deployment

Build docs developers (and LLMs) love

Installation

Basic example

Expected output

What’s happening?

Client types

Next steps