Free · $0, forever
MIT licensed. Self-hosted. Every feature included.
Run itWhat you pay for
- Your server: one container with an SQLite file, on a small machine.
- Your model provider’s tokens, while you build dashboards, alerts and reports, and when someone asks a question or has a panel explained.
What you never pay for
Questions
Does my data leave my infrastructure?
Only to the model provider you choose, and only what the access gate allows. Viewing dashboards involves no model.
Which models work?
Anthropic, OpenAI, Mistral, and any OpenAI-compatible endpoint, such as Gemini, LiteLLM or Ollama.
Can I use a local model?
Yes, through an OpenAI-compatible endpoint such as Ollama.
What does it run on?
One Docker image, for linux/amd64 and linux/arm64.
Is there a hosted version?
No. quanthea runs on your server.