Chroma
Chroma (Chroma DB) is an open-source, AI-native vector database (embedding database) for building LLM, RAG, and semantic-search applications. It provides storage, indexing, and retrieval for vector embeddings with metadata filtering, full-text and regex search, and multi-modal retrieval across text, images, and audio. Chroma is Apache-2.0 licensed and self-hostable via a single server, and is also available as Chroma Cloud, a serverless, usage-based managed offering. Its HTTP/REST v2 API is organized around tenants, databases, collections, and the records (embeddings) inside each collection - add, upsert, update, get, query (similarity search), and delete - and is the same interface used by the Python, JavaScript/TypeScript, Rust, and other client libraries.
APIs
Chroma Collections API
Create, list, get, update, count, and delete collections - the named vector stores that hold embeddings and their metadata. Collections are scoped to a tenant and database and c...
Chroma Records (Embeddings) API
Write and manage the records (embeddings) inside a collection - add, upsert, update, get, count, and delete vectors along with their documents, metadata, and URIs. Records can b...
Chroma Query and Similarity Search API
Run nearest-neighbor vector similarity search over a collection with query embeddings, plus metadata (`where`) and full-text (`where_document`) filters, returning the closest re...
Chroma Tenants and Databases API
Administer the multi-tenancy hierarchy that scopes all collections - create and get tenants, and create, list, get, and delete databases within a tenant. Every collection and re...
Chroma System and Health API
Operational endpoints for a Chroma server - heartbeat, healthcheck, version, and pre-flight checks. On self-hosted deployments a reset endpoint can clear all data when explicitl...