kotaemon
An open-source, customizable RAG chat interface for asking questions over your own documents, aimed at both end users and developers who want to build their own retrieval pipeline.
kotaemon is an open-source, clean and customizable RAG user interface for chatting with documents, built with both end users and developers in mind. For end users it offers a minimalistic interface for question answering over documents, support for LLM API providers such as OpenAI, Azure OpenAI and Cohere as well as local models through Ollama and llama-cpp-python, and simple installation scripts.
For developers it provides a framework for building a custom RAG pipeline, with a customizable UI built on Gradio. Features include multi-user login with private and public document collections, a hybrid full-text and vector retriever with re-ranking, multi-modal question answering over documents with figures and tables, citations with relevance scores shown in an in-browser PDF viewer, question decomposition for complex multi-hop questions, agent-based reasoning such as ReAct and ReWOO, and a settings UI for adjusting retrieval and generation, including prompts.
It is written in Python and licensed Apache-2.0.