Hugging Face Inference Endpoints

United States

Hugging Face Inference Endpoints is a managed model-serving service that provides autoscaling HTTP and gRPC endpoints. It integrates with private Hugging Face Hub repositories and offers enterprise controls including SSO/SCIM, RBAC, audit logs, and VPC connectivity.

From the provider's own site · huggingface.co ↗

Offers tracked10Each links to its source

Accelerators7A100 · A10G · H100

Lowest hourly rate$0.50T4 · AWS · 1 GPU · 14GB each

Regions1Region not listed

Offer list

Every Hugging Face Inference Endpoints offer we track

AcceleratorConfigurationVRAMHourlyRentalAvailabilityT4AWS · 1 GPU · 14GB each14 GB$0.50dedicatedunknownT4GCP · 1 GPU · 16GB each16 GB$0.50dedicatedunknownL4GCP · 1 GPU · 24GB each24 GB$0.70dedicatedunknownL4AWS · 1 GPU · 24GB each24 GB$0.80dedicatedunknownA10GAWS · 1 GPU · 24GB each24 GB$1.00dedicatedunknownL40SAWS · 1 GPU · 48GB each48 GB$1.80dedicatedunknownA100AWS · 1 GPU · 80GB each80 GB$2.50dedicatedunknownA100GCP · 1 GPU · 80GB each80 GB$3.60dedicatedunknownH200AWS · 1 GPU · 141GB each141 GB$5.00dedicatedunknownH100GCP · 1 GPU · 80GB each80 GB$10.00dedicatedunknown