Tiny dependency-free availability prober for any OpenAI-compatible LLM gateway — one short real call per model, 2xx=up. Single static binary, Docker/systemd/Actions deploy.
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).