Minima provides conversational RAG capabilities entirely on-premises, enabling users to query local documents using various LLM providers. It supports fully local deployment with Ollama, custom LLM endpoints via OpenAI-compatible APIs (vLLM, TGI), or cloud integrations with ChatGPT and Anthropic Claude. The system orchestrates vector indexing with Qdrant, embeddings via sentence-transformers, and retrieval pipelines built on LangChain.