Neura Market
Chroma logo

Chroma

Free

Chroma 官方集成，用于嵌入、向量搜索、文档存储和全文搜索。

FreeFree tier
Inputs: text
Type
Open Source
Company
Chroma
Visit Website GitHub (40 stars)

About Chroma

Chroma MCP is a Model Context Protocol (MCP) server implementation that provides database capabilities for Chroma, the open-source embedding database. It enables AI models to create collections over generated data and user inputs, and retrieve data using vector search, full text search, and metadata filtering. The server supports flexible client types (ephemeral in-memory, persistent file-based, HTTP client for self-hosted Chroma, and cloud client for Chroma Cloud). It offers collection management tools (create, modify, delete, list, peek, get info/count), document operations (add, query, get, update, delete), and multiple embedding functions (default, Cohere, OpenAI, Jina, VoyageAI, Roboflow). Users can configure HNSW parameters and use advanced filtering and pagination. Designed for self-hosting, it integrates seamlessly with LLM applications to provide memory and retrieval-augmented generation.

Key Features

Flexible client types: ephemeral, persistent, HTTP client, and Chroma Cloud client
Collection management with pagination, HNSW configuration, and embedding function selection
Document operations: add, query, get, update, delete with metadata and custom IDs
Semantic vector search, full text search, and advanced metadata filtering
Support for multiple embedding functions: default, Cohere, OpenAI, Jina, VoyageAI, Roboflow
Pagination and filtering for listing collections and retrieving documents
Self-hosted or cloud-based deployment via Chroma Cloud

Pros & Cons

Pros
  • Open-source and free to use
  • Supports multiple deployment modes (in-memory, persistent, HTTP, cloud)
  • Integrates with popular embedding models (OpenAI, Cohere, Jina, etc.)
  • Comprehensive collection and document management tools via MCP
  • Easy to self-host or connect to Chroma Cloud
Cons
  • Requires self-hosting for persistent or HTTP client options
  • Dependency on Chroma ecosystem and MCP protocol
  • Limited to Chroma's supported embedding functions and configuration

Best For

Adding memory and retrieval to LLM applicationsBuilding RAG (Retrieval-Augmented Generation) systems with vector searchManaging structured and unstructured document collectionsPrototyping with in-memory Chroma, then scaling to persistent or cloud storagePowering AI-driven search and recommendation features

FAQ

What is Chroma MCP?
Chroma MCP is a Model Context Protocol server implementation that provides database capabilities for Chroma, the open-source embedding database. It allows AI models to create and query collections using vector search, full text search, and metadata filtering.
What client types does Chroma MCP support?
It supports ephemeral (in-memory) for testing, persistent for file-based storage, HTTP client for self-hosted Chroma instances, and cloud client for Chroma Cloud integration.
Can I use Chroma MCP with Chroma Cloud?
Yes, the cloud client automatically connects to api.trychroma.com for Chroma Cloud integration.
What embedding functions are supported?
Chroma MCP supports default, Cohere, OpenAI, Jina, VoyageAI, and Roboflow embedding functions.
Is Chroma MCP free?
Yes, Chroma MCP is open-source and free to use. The tool is released under an open-source license on GitHub.