PrivateGPT

What is PrivateGPT?

PrivateGPT is an open-source project from Zylon that lets developers build AI applications whose data never leaves their own infrastructure. It began in 2023 as an offline document-question pipeline, and version 1.0 reworks it into an application API that sits between your app and a self-hosted inference server such as Ollama, vLLM, llama.cpp or LM Studio. It does not run models itself. Instead it supplies the pieces a cloud AI API normally provides: a messages endpoint, file and document ingestion, retrieval with citations, custom tools, MCP connectors, database and CSV access, web search, code execution, embeddings and token counting. Because it follows the Claude API specification, clients written for Claude, including Claude Code and the Office add-ins, can be pointed at it. A lightweight UI is included for testing. The code is free to self-host; the same team sells Zylon, an on-premise enterprise platform whose prices are not published, with demos on request.

Use cases

• Build private RAG apps on local models
• Chat with documents without sending data out
• Run Claude-compatible clients on self-hosted LLMs
• Add tool use and MCP to local inference
• Query databases and CSVs privately
• Prototype air-gapped AI assistants

Who it's for

• Developers
• Enterprise IT teams
• Regulated-industry teams
• AI engineers

Why use PrivateGPT

• Documents, prompts and answers stay on your own servers, which helps teams that cannot send data to external AI APIs.
• You keep your choice of local model server and swap it without rewriting the app layer.
• Tools built for the Claude API, such as Claude Code, can run against your own infrastructure.
• Local inference avoids per-token cloud fees as usage grows.
• The project has a large public following, with 57k stars and 8k+ forks on GitHub.

PrivateGPT features

• Messages API with streaming, async calls and token counting
• File and document ingestion
• Retrieval with citations and agentic RAG
• Custom tool definitions and MCP connectors
• Web search, web fetch and code execution tools
• Structured access to databases and CSV files
• Embeddings endpoint
• Compatibility with the Claude API specification
• Works above Ollama, vLLM, llama.cpp or LM Studio
• Lightweight UI for testing

PrivateGPT integrations

• Ollama
• vLLM
• llama.cpp
• LM Studio
• Claude Code
• n8n
• OpenCode
• Qdrant
• LlamaIndex

How to get started with PrivateGPT

1. Run an inference server such as Ollama or vLLM that exposes OpenAI-compatible endpoints.
2. Deploy PrivateGPT and point it at that server.
3. Call its messages API from your app, agent or workflow, uploading files for retrieval.
4. Add custom tools or MCP connectors where the app needs them.
5. Use the built-in UI to test, then connect your own front end or Claude-compatible client.

PrivateGPT limitations

• Does not run models; you need a separate OpenAI-compatible inference server
• Self-hosted, so you operate and maintain the infrastructure
• The bundled UI is meant for testing, not as a finished product
• Zylon enterprise pricing is not published

PrivateGPT pricing

PrivateGPT is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

PrivateGPT alternatives

The main PrivateGPT alternatives listed on Debuly are AnythingLLM, Open WebUI, Dify, Ollama, LM Studio and LocalAI. The table compares pricing and platforms; order is A–Z.

1. 

   AnythingLLM

   Open-source AI app for chatting with your documents and running agents, on desktop, in Docker or as a hosted team workspace.

   Chat assistantsAutomation and agentsDeveloper platforms
   Open sourceVisit AnythingLLM (official site)
2. 

   Dify

   Platform for building LLM apps, agents and RAG workflows on Dify Cloud or your own infrastructure, with a free Sandbox tier.

   Developer platformsAutomation and agentsChat assistants
   FreemiumVisit Dify (official site)
3. 

   LM Studio

   Desktop app for running open language models on your own computer, now with the Bionic agent, a local API and optional US-hosted cloud models.

   Chat assistantsDeveloper platformsCodingProductivity
   FreemiumVisit LM Studio (official site)
4. 

   LocalAI

   MIT-licensed open-source runtime that serves local models through OpenAI- and Anthropic-compatible APIs for text, images, audio and agents.

   Developer platformsAutomation and agentsChat assistants
   Open sourceVisit LocalAI (official site)
5. 

   Ollama

   Open-source app and CLI for running open AI models locally on macOS, Windows and Linux, with optional paid cloud models.

   Developer platformsCodingChat assistants
   Open sourceVisit Ollama (official site)
6. 

   Open WebUI

   Self-hosted AI chat platform for Ollama and OpenAI-compatible models with RAG, agents, tools, notes, channels and role-based access.

   Chat assistantsDeveloper platformsAutomation and agents
   Open sourceVisit Open WebUI (official site)

PrivateGPT FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from framerusercontent.com · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.