Cerebrium vs Deep Infra
As of , Deep Infra lists the lower on-demand rate on 4 of the 4 GPUs both offer on-demand. On the priciest shared card both list on-demand, NVIDIA B200: Cerebrium $6.01 vs Deep Infra $4.89 per GPU-hour on-demand — Deep Infra is $1.12 (19%) cheaper.
On-demand, per GPU-hour
| GPU | Difference | Cheaper | ||
|---|---|---|---|---|
| NVIDIA B200 median $6.94 · 40 providers · lowest Hostinger $4.50 from | $6.01 GPU only captured Deploy ↗ | $4.89 captured Deploy ↗ | $1.12 (19%) | Deep Infra |
| NVIDIA H200 median $4.56 · 56 providers · lowest Dataoorts $2.29 | $4.20 GPU only captured Deploy ↗ | $3.49 captured Deploy ↗ | $0.71 (17%) | Deep Infra |
| NVIDIA H100 80GB (form factor not stated) median $3.75 · 28 providers · lowest Gcore $1.72 | $3.40 GPU only captured Deploy ↗ | $2.90 captured Deploy ↗ | $0.50 (15%) | Deep Infra |
| NVIDIA A100 80GB median $1.99 · 63 providers · lowest Yotta Labs $0.92 | $2.10 GPU only captured Deploy ↗ | $1.49 captured Deploy ↗ | $0.61 (29%) | Deep Infra |
How this is compared
Cloud GPU rental prices of Cerebrium and Deep Infra, 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.
GPU by GPU
- NVIDIA B200: Cerebrium $6.01 vs Deep Infra $4.89 per GPU-hour on-demand — Deep Infra is $1.12 (19%) cheaper.
- NVIDIA H200: Cerebrium $4.20 vs Deep Infra $3.49 per GPU-hour on-demand — Deep Infra is $0.71 (17%) cheaper.
- NVIDIA H100 80GB (form factor not stated): Cerebrium $3.40 vs Deep Infra $2.90 per GPU-hour on-demand — Deep Infra is $0.50 (15%) cheaper.
- NVIDIA A100 80GB: Cerebrium $2.10 vs Deep Infra $1.49 per GPU-hour on-demand — Deep Infra is $0.61 (29%) 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: