L4 pricing on Hugging Face
Every Hugging Face L4 offering we track, normalized to a per-GPU hourly rate and ranked against 22 providers with current L4 pricing.
The cheapest L4 we currently track is $0.321/hr on Vast.ai.
Hugging Face L4 rates
Hourly prices per GPU. Multi-GPU rows show the per-GPU rate, not the instance total.
| Price / GPU-hr | Type | GPUs |
|---|---|---|
| $0.700 | On-demand | 1 |
| $0.800 | On-demand | 1 |
| $0.950 | On-demand | 4 |
| $0.950 | On-demand | 4 |
L4 specifications
- VRAM
- 24GB
- Architecture
- Ada Lovelace
- CUDA Cores
- 7,424
- Tensor Cores
- 232
- Memory Bandwidth
- 300 GB/s
- TDP
- 72W
L4 on other providers
- Vast.ai$0.321/hr
- Omega Gradient$0.360/hr
- Seeweb$0.433/hr
- gpu.ai$0.440/hr
- Jarvis Labs$0.440/hr
- AtmosCompute$0.450/hr
- Runpod$0.490/hr
- Aquanode$0.539/hr
- Google Cloud$0.600/hr
- UpCloud$0.650/hr
- AceCloud$0.662/hr
- Modal$0.799/hr
- Amazon AWS$0.805/hr
- Baseten$0.848/hr
- Scaleway$0.898/hr
- Shadeform$0.950/hr
- OVHcloud$1.00/hr
- IO.NET$1.04/hr
- Sesterce$1.04/hr
- Runcrate$1.05/hr
- Hinode$1.75/hr
About Hugging Face
Hugging Face runs the Hub for open models and datasets. Inference Endpoints deploys a model on dedicated GPUs on AWS or Google Cloud, billed per hour of instance time, and Inference Providers routes API requests to partner inference providers at those providers' own rates.
Frequently Asked Questions
How much does the L4 cost on Hugging Face?
Hugging Face lists the L4 from $0.700 per GPU-hour across 4 tracked configurations. Prices are collected daily — see the table above for the current rates.
Is Hugging Face the cheapest place to rent the L4?
Hugging Face ranks #12 of 22 providers we track with current L4 pricing. The cheapest right now is Vast.ai. Price is only one factor — region availability, network, and storage costs differ between providers.
Which L4 configurations does Hugging Face offer?
We currently track 4 L4 offerings from Hugging Face, covering the GPU counts, pricing types, and regions listed in the table above. Every price is normalized to a per-GPU hourly rate so it can be compared across providers.