Skip to content

Minima

An open-source local RAG for querying your own files

Visit official site

What Minima is

The model is described as an open-source, fully local RAG (Retrieval-Augmented Generation) model that integrates with ChatGPT and MCP (Model Context Protocol), processing data locally to protect privacy.

What you can do with it

  • Install fully locally with no external network connection
  • Query local files through ChatGPT and Anthropic Claude
  • Use custom GPT integration for local document retrieval
  • Deploy with Docker containers and configure environment variables

Application scenarios

  • A company builds an internal knowledge base for staff search
  • A developer creates a custom document retrieval tool for a project
  • An educational institution integrates teaching materials for student search

Getting started

  • Visit the Minima GitHub page and clone the project locally
  • Create a .env file and configure the relevant environment variables
  • Choose a Docker Compose file that matches the use case
  • Run the corresponding Docker Compose command to start the services

Pros & cons

✓ What we like

  • Free and open source
  • Local processing protects data privacy

! What to watch out for

  • Modes are tied to ChatGPT, Anthropic Claude or a fully local install
  • Setup involves Docker Compose files and environment variables

FAQ

What are the three modes?

A fully local install, querying local documents through ChatGPT, and querying local files with Anthropic Claude.

Which file formats are named?

The feature list names .pdf, .xls, .docx, .txt, .md and .csv.

Last reviewed: 2026-11-15

More AI coding tools tools

View all →

How we review