Which Scaleway GPU is best for LLM Inference?

Based on AIMC's fit-scoring methodology, the highest-fit Scaleway GPU for LLM Inference is the L40S at $1.59/hr (fit score 100/100, excellent fit). Total of 5 viable GPUs tracked at Scaleway. AIMC has no commercial relationship with Scaleway.

What is the cheapest Scaleway GPU suitable for LLM Inference?

The cheapest Scaleway GPU with an adequate-or-better fit for LLM Inference is the L4 at $0.85/hr (fit score 90/100). Reference pricing sourced from publicly listed catalogs.

Hyperscaler Reference/Scaleway/for LLM Inference

Workload Analysis · Reference Pricing

Best Scaleway GPUs
for LLM Inference

Fit-scoring analysis of 5 viable Scaleway GPUs for LLM Inference. Cheapest from $0.85/hr. Highest fit: L40S (100/100).

AIMC's Picks for LLM Inference

Selected from 5 viable Scaleway GPUs using AIMC's fit-scoring methodology.

Best Overall Fit

L40S

Highest fit-score for this workload

$1.59/hrExcellent fit · 100/100

Best Value

Cheapest with adequate-or-better fit

$0.85/hrExcellent fit · 90/100

Best for Production

H100 PCIe

Highest-fit datacenter-class GPU

$3.10/hrExcellent fit · 100/100

All Viable Scaleway GPUs for LLM Inference

Sorted by fit score descending. Click any GPU for full reference pricing detail.

Datacenter GPUs100/100

H100 SXM

REF

$3.42/hr

Datacenter GPUs100/100

B300 SXM

REF

$8.10/hr

Datacenter GPUs100/100

REF

$0.85/hr

Datacenter GPUs90/100

Other Workloads at Scaleway

3D Rendering5 GPUs · fit 100

Audio AI5 GPUs · fit 100

Computer Vision5 GPUs · fit 100

Embedding Generation5 GPUs · fit 100

Other Hyperscalers for LLM Inference

Legal Notice — Independent Market Intelligence

AIMC has no commercial relationship or data-sharing agreement with Scaleway. Pricing data is sourced from publicly listed catalogs. "Scaleway", "Scaleway", and related marks are trademarks of their respective owners. Fit-scoring is AIMC's independent analysis based on published GPU specifications.