Deployment
Model deployment
Run compatible models without building and operating a separate GPU platform for every provider.
What QDivZero manages
QDivZero validates model requirements, searches compatible capacity across providers, prepares the runtime, mounts persistent storage and operates the deployment lifecycle.
- Models from Hugging Face and other supported sources.
- Capacity selection across compatible providers.
- Persistent storage independent of a single provider.
- Stable serving names and OpenAI-compatible access.
Capacity pricing
Deployments use active-capacity pricing. The selected per-second rate is fixed when the deployment starts and usage is recorded by active time, not by the number of tokens processed.
Operate from one place
Start, stop, inspect and schedule deployments from QDivZero. Add LLM Firewall or place deployments behind LLM Model Routing without changing the client contract.