HF Endpoints
Hugging Face hosts the largest public library of open models. Its Inference Endpoints run a chosen model on dedicated GPUs by the hour, on AWS or Google Cloud.
- Website: https://huggingface.co
- Headquarters: New York, United States
- Kind: Model hub, Dedicated endpoints
HF Endpoints sells 0 of the compared models per token and prices 7 kinds of GPU for hosting your own.
GPUs for hosting your own
| GPU | From, per GPU-hour | Cheapest instance | GPUs in it |
|---|---|---|---|
| T4 | $0.50 | aws.nvidia-t4.x1 | 1 × 14 GB |
| L4 | $0.70 | gcp.nvidia-l4.x1 | 1 × 24 GB |
| A10G | $1.00 | aws.nvidia-a10g.x1 | 1 × 24 GB |
| L40S | $1.80 | aws.nvidia-l40s.x1 | 1 × 48 GB |
| A100 | $2.50 | aws.nvidia-a100.x1 | 1 × 80 GB |
| H200 | $5.00 | aws.nvidia-h200.x1 | 1 × 141 GB |
| H100 | $10.00 | gcp.nvidia-h100.x1 | 1 × 80 GB |
Where it runs
One global price; HF Endpoints chooses where a request is processed.
Where the prices come from
Sources
List prices collected 4 November 2026, refreshed every three hours. No negotiated rates or taxes.