privateGPT
Open-source tool that lets you chat with your documents privately using local language models, with no data sent to external servers.
About
privateGPT is an open-source API layer that enables fully offline, private document chat using locally running language models. You point it at your files and ask questions in plain language; it retrieves relevant passages and generates answers without any data leaving your machine or network, making it suitable for air-gapped and compliance-sensitive environments.
It supports any OpenAI-compatible inference server, connects to tools like Notion, JIRA, and Slack with role-based permissions, and provides capabilities including document ingestion, retrieval with citations, tabular data analysis, and MCP tool integration. The open-source version is free to self-host; Zylon offers a supported enterprise on-premise product built on top of it.
It supports any OpenAI-compatible inference server, connects to tools like Notion, JIRA, and Slack with role-based permissions, and provides capabilities including document ingestion, retrieval with citations, tabular data analysis, and MCP tool integration. The open-source version is free to self-host; Zylon offers a supported enterprise on-premise product built on top of it.