Inference Endpoint instance hour
Catalog row in Vendor License Metrics · Cited
- Unit
- Hour of instance per running replica (billed per minute)
- Definition
- Cost is "instance hourly rate * ((hours * # min replica ) + (scale-up hrs * # additional replicas))" according to the pricing examples.
- Counting rules
- Charged while a successfully deployed Endpoint is initializing or running. Scaled-to-zero Endpoints count as used quota until paused.
- Floors / minimums
- Instance types may require a quota request.
- Applies to products
- Inference Endpoints
- Related metrics
- Product category