private-gpt

private-gpt

Provides a complete API layer for local models, supporting RAG, skills, tools, and more.

MCPDevOpen source
Type
MCP
Transport
http
Open source
Yes
GitHub Stars
★ 57.3k
Source
mcp-github

Overview

PrivateGPT is an open-source API layer that transforms local models into production-grade AI applications. It supports various features such as standard message APIs, file and artifact ingestion, retrieval citations, built-in tools (e.g., web search, code execution), custom tools and MCP connectors, structured access to databases and CSVs, as well as embeddings and orchestration. By connecting to any OpenAI-compatible inference server, PrivateGPT enables developers to build private AI products without rebuilding backend components from scratch. Ideal for enterprises and individuals who need to run AI applications locally.

Capabilities

  • Standard message API
  • File and artifact ingestion
  • Retrieval citations
  • Built-in tools (e.g., web search)
  • Custom tools and MCP connectors
  • Structured access to databases and CSVs

Use cases

Local AI application developmentEnterprise internal AI platformsPrivate data processingCode execution and debugging

Setup

Requires: OpenAI兼容的LLM服务器Node环境
brew tap zylon-ai/tap && brew install private-gpt or install using the curl script

This information was compiled by AI from public sources and may contain inaccuracies — please refer to the source.

FAQ

Does PrivateGPT run the model?

No, it connects to any OpenAI-compatible inference server.

How do I start PrivateGPT?

Set environment variables and run the `private-gpt serve` command.

Related skills