Chroma

关于

Chroma MCP Server acts as an enhanced bridge between Large Language Model (LLM) applications and external data sources using Chroma, the open-source embedding database. This specialized fork of the original chroma-mcp project offers a FastMCP architecture, streamable HTTP transport, and improved production readiness. It enables AI models to create and manage collections of generated data and user inputs, facilitating powerful data retrieval through vector search, full-text search, and metadata filtering, providing LLMs with the essential context they need.

主要功能

• FastMCP Architecture with streamable HTTP transport
• Flexible client types: ephemeral, persistent, HTTP, and Chroma Cloud
• Comprehensive collection management (create, modify, delete, query)
• Document operations including add, query (semantic, full-text, metadata), retrieve, update, and delete
• Support for multiple embedding functions (default, Cohere, OpenAI, Jina, VoyageAI, Roboflow)
• 0 GitHub stars

使用案例

• Providing LLMs with memory by storing and retrieving conversational context and generated data
• Integrating external data sources and tools with LLM applications through a standardized protocol
• Building Python or JavaScript LLM applications that require vector search and data retrieval capabilities