Self-host the gateway
Run an OpenAI-compatible HAL Model Gateway on your own machine or VPS. Cloud-first by default — no local Ollama required inside the image.
Why self-host
- Your API keys and gateway token stay on your host — never baked into the image.
- Same
/v1/modelsand/v1/chat/completionssurface as the public API. - Optional Piper voices via volume; chat works without TTS.
Docker one-liner
git clone https://github.com/UniteAndCreateForLife/hal-supreme-community.git cd hal-supreme-community/docker/model-gateway python3 prepare-context.py cp .env.example .env mkdir -p secrets && openssl rand -base64 32 > secrets/hal_gateway_token && chmod 600 secrets/hal_gateway_token # edit .env — add at least GROQ_API_KEY (or another remote) docker compose up --build -d curl -fsS http://127.0.0.1:8767/health
Port 8767 on the host. Auth via HAL_GATEWAY_TOKEN or the secret file. Never paste tokens into chat, Issues, or Discord.
Docs & community
© 2026 HAL SUPREME · Quantum Black / HAL Blue