GPU Marketplace
Estimates only — rates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .
Company Profile
Legal EntityScaleway
Infrastructure
GPU FleetL40S, L4, H100 PCIe, H100-SXM, B300-SXM, (also Titanium (GPU) and GPU Dedibox in Bare Metal listings)
Bare MetalBare Metal (Elastic Metal, Dedibox) available
Compute & Deployment
On-DemandPrice (€/hour)
Bare MetalElastic Metal
VM-BasedVirtual Instances
Container-BasedContainers
KubernetesKubernetes Kapsule
GPU Hardware
Latest GenL40S, L4, H100 PCIe, H100-SXM, B300-SXM
Multi-GPU NodesH100 GPU instances support multi-GPU (up to 8 GPUs per instance)
Max GPUs/Node8
PCIe vs SXMH100 PCIe and H100-SXM (also B300-SXM)
HGX PlatformHGX B300 (Blackwell architecture)
Pricing Model
Public Pricinghttps://www.scaleway.com/en/pricing/gpu/
Hidden FeesFlexible IP is billed at €0.004/hour.
Performance & Scaling
Developer Experience
Security & Compliance
Data Center Locations
Coverage
CitiesParis, Amsterdam, Warsaw, Milan
ParisAmsterdamWarsawMilan
Compliance Regions
Datacenter Locations
Key Strengths
Offers multiple GPU instance families and bare-metal GPU options geared for AI/accelerated workloads
L4 described as flexible and cost-effective for AI
L40S positioned as universal GPU instance "faster than L4 and cheaper than H100 PCIe"
H100 instances support up to 8 GPUs per instance
B300-SXM (HGX B300) offers very large memory (up to 2.3 TB) for high-end AI reasoning.
Known Limitations
Pricing shown as hourly and monthly approximations only (no spot pricing option shown)
some GPU types are not available in all regions/AZs (availability tables show Not available for many regions/AZs).
Additional Information
Last updated August 2026. Information subject to change.