Our take
Quietno commits in the last 90 days- Activity
- 36
- Community
- 71
- Issues
- 21
- Pull requests
- 20
+103 stars in the last 30 days
kotaemon is a feature-rich starting point for self-hosted document QA, with a quiet recent development period but several releases in the past year and contributions from a broad group.
Good fit if
- You want a self-hosted RAG web app with citations, document collections, and configurable retrieval.
- You are a Python developer looking to adapt a Gradio-based RAG interface or pipeline.
- Apache-2.0 licensing suits your deployment and modification plans.
Look elsewhere if
- You need evidence of frequent recent commits or issue and pull request throughput, since none were recorded in the last 90 days.
- You need a turnkey document-processing setup for file types beyond PDF, HTML, MHTML, and XLSX without installing additional packages.
- Your project requires a fully documented custom indexing workflow, which the README describes as still in progress.
All health signals
| Last commit | 2026-07-14 (3 months ago) |
|---|---|
| Commits, last 90 days | 0 |
| Releases, last 12 months | 4 (latest v0.12.0, 2026-05-31) |
| Contributors | 39 (top contributor: 37% of commits) |
| Issues closed, last 90 days | 0 |
| Pull requests merged, last 90 days | 0 |
| Project age | 2 years |
Checked on 2026-10-24 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
25,912Forks
2,168Change since Nov 16, 2026: +2 stars·+1 forks
Overview
kotaemon is an open-source RAG interface and framework built with Gradio, the UI library from Gradio. It gives people a web app for asking questions about their documents, while also letting developers adapt the underlying pipeline instead of building the interface and document QA flow from scratch.
It is aimed at both individuals or teams who want to host document Q&A themselves and developers who need a customizable base for RAG applications. It supports local deployment with Python or Docker, and can connect to hosted LLM providers or local models.
Key Features
- Web UI for document question answering, with multi-user accounts and private or public collections.
- Hybrid retrieval combining full-text and vector search, with reranking.
- Citations with relevance scores, PDF previews, and highlighted passages.
- Document QA for files containing figures and tables, with selectable parsing integrations.
- Configurable LLMs, embedding models, retrieval settings, prompts, and reasoning pipelines.
- Support for question decomposition and agent-based reasoning methods such as ReAct and ReWOO.
- Extensible indexing and retrieval, with GraphRAG implementations provided as examples.
Use Cases
- A team can self-host a document library and let members ask questions with citations and PDF previews.
- Developers can use the provided Gradio app to prototype a RAG pipeline, then customize its UI and reasoning components.
- People working with private documents can configure local models through Ollama or llama-cpp-python rather than relying only on hosted model APIs.
- Users handling PDFs with figures or tables can configure multimodal parsing integrations for document QA.
What you need
Detected in the repository
- Python >= 3.10 (from pyproject.toml)
- A Dockerfile, so it can run in a container
- Automated checks on GitHub Actions
License in plain words
Apache-2.0permissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the license, the NOTICE file and a note of your changes
- Share your changes: no
- Includes an explicit patent grant from the contributors:
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
With Docker, run the lightweight image and open http://localhost:7860/:
docker run -e GRADIO_SERVER_NAME=0.0.0.0 -e GRADIO_SERVER_PORT=7860 -v ./ktem_app_data:/app/ktem_app_data -p 7860:7860 -it --rm ghcr.io/cinnamon/kotaemon:main-lite
See the README and user guide for setup details, model configuration, and other installation options.
Alternatives
- Article-Assistant--RAG-Telegram-Bot: It offers document and article Q&A through a Telegram bot, rather than a customizable self-hosted Gradio interface and RAG framework.
- RAG-Anything: It focuses on multimodal document ingestion and retrieval built on LightRAG, rather than providing a customizable Gradio interface and framework.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| kotaemon | Python | Apache-2.0 | 25.8k | Quiet |
| Article-Assistant--RAG-Telegram-Bot | Python | MIT | 54 | Inactive |
| RAG-Anything | Python | MIT | 23.5k | Active |
Considerations
The project is marked quiet: it has had no commits in the last 90 days, although it has continued to publish releases during the past year. Its 39 contributors are not highly concentrated around one person, but the measured data shows no recent issue closures or merged pull requests, so teams should assess their support and maintenance expectations. Python 3.10 or newer is required; Docker is optional. Processing some document formats beyond PDF, HTML, MHTML, and XLSX requires installing Unstructured, and local-model operation requires separately setting up a supported runtime such as Ollama or llama-cpp-python. The README also flags the custom indexing instructions as work in progress.
Found this useful?
Share it with someone who would like kotaemon.