PrivateGPT

An API layer for building private AI applications on local models, with RAG, tools, MCP and text-to-SQL, connecting to any OpenAI-compatible model server.

57.6k stars, 7.6k forks Python On GitHub since May 2023

Visit website View on GitHub How to run it

What is PrivateGPT?

PrivateGPT provides the backend pieces that a private AI product needs on top of a local model: document ingestion and retrieval, tools, skills, MCP support and text-to-SQL. It does not run models itself; it connects to any inference server that implements the OpenAI chat completions endpoint, and its own API follows the shape of the Claude API. A small workbench UI is included for testing and demos. It also powers Zylon, an on-premises AI platform sold to enterprises by the same team.

Good fit if

  • You are building an internal AI product and want ready RAG and tool calling without cloud APIs.
  • You already run a model server such as vLLM, llama.cpp or Ollama.

Look elsewhere if

Before you adopt it

PrivateGPT is Apache-2.0, with no paid parts in this repository; Zylon is the separate commercial product. Docker images exist, but you must bring your own model server.

Drafted with AI from the project's README and repository data, then checked against that data on 30 October 2026. Report a correction

PrivateGPT health: is it still maintained?

67 of 100, actively developed

  • Activity
    55 commits in the last 90 days, the latest on 29 October 2026.
    21 / 30
  • Contributors
    16 people wrote the 55 commits of the last 90 days; the most active wrote 48% of them.
    17 / 20
  • Releases
    2 releases in the last 12 months; the latest, v1.0.1, on 18 June 2026.
    11 / 20
  • Replies to issues
    Maintainers answered 7% of recent issues, with a median first reply after 5 hours.
    13 / 20
  • Security practices
    No OpenSSF Scorecard for this repository yet, so this part gets a neutral score.
    5 / 10

Read from GitHub on 1 November 2026. How the score works

PrivateGPT license: is it open source?

PrivateGPT uses Apache-2.0, an open-source license approved by the OSI.

  1. May 2023First published on GitHub.
  2. TodayApache-2.0.

How to run PrivateGPT yourself

Docker
Dockerfile or compose file in the repository
Compose file
None in the repository
Setup effort
Not rated: no compose file to judge from
Kubernetes
No Helm chart in the repository
Desktop and mobile
No app builds in GitHub releases
Interface languages
English only

Install instructions: zylon.ai or the README on GitHub.

Alternatives to PrivateGPT

  • RAGFlowRAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG… 93 health · Apache-2.0 · 93.3k stars
  • DocsGPTPrivate AI platform for agents, assistants and enterprise search. Built-in Agent Builder, Deep research,… 84 health · MIT · 18.4k stars
  • HarborStop configuring your AI stack. Start using it. One command brings a complete pre-wired LLM stack with… 80 health · Apache-2.0 · 3.3k stars
  • OnyxOpen Source AI Platform - AI Chat with advanced features that works with every LLM 77 health · MIT · 33.1k stars
  • LibreChatEnhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API,… 74 health · MIT · 46.6k stars

All 36 projects in AI chat & local LLMs

Questions about PrivateGPT

Is PrivateGPT open source?

Yes. PrivateGPT is published under Apache-2.0, an OSI-approved open-source license.

Is PrivateGPT still maintained?

Its health score is 67 of 100 (actively developed). It had 55 commits in the last 90 days, the latest on 29 October 2026, and its latest release came out on 18 June 2026.

How do I run PrivateGPT?

You can run it with Docker. The README on GitHub has the exact steps.

What are the alternatives to PrivateGPT?

Close alternatives in the AI chat & local LLMs category include RAGFlow, DocsGPT, Harbor.