Chroma

Overview

Chroma is an open-source search infrastructure for AI. It offers fast, serverless, and scalable infrastructure supporting vector, full-text, regex, and metadata search, built on object storage and used by millions of developers.

Key Features

• Vector search for semantic similarity search.
• Sparse vector search for lexical search with BM25 and SPLADE.
• Full-text search with trigram and regex search.
• Metadata search with filtering and faceted search.
• Forking for dataset versioning, A/B testing, and roll-outs.
• Command-line tools for development.

Use Cases

• AI applications and agents that need search, retrieval, and context engineering.
• Knowledge bases and question-answering workflows.
• Multi-tenant search over billions of indexes.
• Dataset versioning, A/B testing, and roll-outs using forking.

Getting Started

• Install the Python client with pip install chromadb.
• Install the JavaScript/TypeScript client with npm install chromadb.
• Start client-server mode with chroma run --path /chroma_db_path.
• Start free on Chroma Cloud with $5 in free credits, or read the docs.
• Get up and running in 30 seconds or less.

Deployment & Requirements

• Run Chroma OSS on your own infrastructure using open-source deployment guides.
• Use client-server mode with chroma run --path /chroma_db_path.
• Chroma Cloud provides a hosted serverless deployment option.
• Enterprise deployment supports BYOC in your VPC, multi-cloud/multi-region replication, and point-in-time recovery.
• Built on object storage with automatic data tiering and caching.

Before You Adopt

• License: Apache-2.0. Review its terms before using, modifying, or distributing the project.
• Chroma Cloud became generally available in August 2025.
• The project is described as rapidly developing and welcomes contributions.
• Chroma Cloud offers customer-managed encryption keys to encrypt data with users' own keys.
• Chroma Cloud supports private networking through AWS PrivateLink.