FlexInfer docs
FlexInfer specs
CRDs, HTTP contracts, labels, and operational interfaces.
Specs
This section documents the “contracts” FlexInfer exposes:
- Kubernetes CRDs (
Model,ModelDeployment,ModelCache,GPUGroup) - Proxy endpoints and routing behavior
- Scheduler extender endpoints (Kube scheduler extender v1)
- Labels/annotations used for discovery and routing
- Prometheus metrics emitted by components
Entry points:
docs/specs/crds.mddocs/specs/proxy-api.mddocs/specs/scheduler-extender.md