Hugging Face Source

Inference Endpoints pricing documentation

Catalog row in Vendor License Sources · Cited

Kind: Documentation
Format: Web page
Applies to: Dedicated Inference Endpoints instance types, hourly rates by cloud provider, cost formula and quotas. Undated page.
Evidence: Inference Endpoints are billed per minute while initializing or running
Endpoint cost multiplies hourly rate by replicas and scale-up hours
Scaled-to-zero Endpoints still count against instance quota