CoreWeave vs Hugging Face Inference Endpoints
As of , Hugging Face Inference Endpoints lists the lower on-demand rate on 3 of the 5 GPUs both offer on-demand; CoreWeave wins on 2. On the priciest shared card both list on-demand, NVIDIA B200: CoreWeave $8.60 vs Hugging Face Inference Endpoints $9.25 (aws) per GPU-hour on-demand — CoreWeave is $0.65 (7%) cheaper.
Cloud GPU rental prices of CoreWeave and Hugging Face Inference Endpoints, compared per GPU-hour on the GPUs both list. Where a provider has several tiers or regions, its cheapest on-demand rate is used and labelled. Community capacity (RunPod Community Cloud) and marketplace floors are shown only where a provider has nothing else, and never decide which is cheaper.
On-demand, per GPU-hour
| GPU | CoreWeave | Hugging Face Inference Endpoints | Difference | Cheaper |
|---|---|---|---|---|
| NVIDIA B200 | $8.60 captured Deploy ↗ | $9.25 aws captured Deploy ↗ | $0.65 (7%) | CoreWeave |
| NVIDIA H200 | $6.30 captured Deploy ↗ | $5.00 aws captured Deploy ↗ | $1.30 (21%) | Hugging Face Inference Endpoints |
| NVIDIA A100 80GB | $2.70 captured Deploy ↗ | $2.50 aws captured Deploy ↗ | $0.20 (7%) | Hugging Face Inference Endpoints |
| NVIDIA RTX PRO 6000 Blackwell | $2.50 captured Deploy ↗ | $2.75 aws captured Deploy ↗ | $0.25 (9%) | CoreWeave |
| NVIDIA L40S | $2.25 captured Deploy ↗ | $1.80 aws captured Deploy ↗ | $0.45 (20%) | Hugging Face Inference Endpoints |
Spot / interruptible, per GPU-hour
| GPU | CoreWeave | Hugging Face Inference Endpoints |
|---|---|---|
| NVIDIA B200 | $4.26 captured | — |
| NVIDIA H200 | $2.58 captured | — |
| NVIDIA A100 80GB | $1.19 captured | — |
| NVIDIA RTX PRO 6000 Blackwell | $1.20 captured | — |
| NVIDIA L40S | $0.98 captured | — |
GPU by GPU
- NVIDIA B200: CoreWeave $8.60 vs Hugging Face Inference Endpoints $9.25 (aws) per GPU-hour on-demand — CoreWeave is $0.65 (7%) cheaper.
- NVIDIA H200: CoreWeave $6.30 vs Hugging Face Inference Endpoints $5.00 (aws) per GPU-hour on-demand — Hugging Face Inference Endpoints is $1.30 (21%) cheaper.
- NVIDIA A100 80GB: CoreWeave $2.70 vs Hugging Face Inference Endpoints $2.50 (aws) per GPU-hour on-demand — Hugging Face Inference Endpoints is $0.20 (7%) cheaper.
- NVIDIA RTX PRO 6000 Blackwell: CoreWeave $2.50 vs Hugging Face Inference Endpoints $2.75 (aws) per GPU-hour on-demand — CoreWeave is $0.25 (9%) cheaper.
- NVIDIA L40S: CoreWeave $2.25 vs Hugging Face Inference Endpoints $1.80 (aws) per GPU-hour on-demand — Hugging Face Inference Endpoints is $0.45 (20%) cheaper.
Not compared: GPUs only one of the two lists, reserved/committed tiers, and anything either provider prices only on request. See each provider's page for its full list.
Data updated: