Minima
Createddmayboroda
Enables on-premises conversational retrieval augmented generation (RAG) with configurable containers and optional ChatGPT or Claude integration.
About
Minima provides a flexible, open-source solution for on-premises RAG, offering three operational modes: fully isolated operation with local LLMs, integration with ChatGPT for querying local documents via custom GPTs, and integration with Anthropic Claude. It leverages Docker containers for easy deployment and configuration, supporting a range of embedding models, LLMs (via Ollama), and rerankers. With Minima, users can maintain data security and control while leveraging the power of conversational AI.
Key Features
- Integrates with ChatGPT and Anthropic Claude for enhanced conversational capabilities.
- Includes a local electron app for chat UI using a fully local installation.
- Supports fully local RAG operations, ensuring data privacy.
- Supports multiple embedding models, LLMs (via Ollama), and rerankers.
- 664 GitHub stars
- Uses Docker containers for easy deployment and management.
Use Cases
- Deploying a fully isolated, on-premises RAG system for secure data access.
- Using Anthropic Claude to interact with indexed local data.
- Creating a custom GPT that can answer questions based on local documents.