GPU Cloud Provider · Not specified
Crusoe
Crusoe provides advanced cloud solutions with a focus on AI workload optimization, offering managed services, highly reliable infrastructure, and AI-optimized data centers. Their cloud platform is built around NVIDIA and AMD GPUs, boasting optimized orchestration for intelligent workload management and high availability.
GPUs
1
Founded
Not specified
Countries
1
Data Centers
4
Team Size
201-1000
GPU Marketplace
NVIDIA A100 80GB SXMOn-Demand
Company Profile
Company TypeScale-up
Provider TypeCloud Provider
FoundedNot specified
HeadquartersNot specified
Legal EntityCrusoe Energy Systems, Inc.
FundingSeries D
Total Raised$686M
Team Size201-1000
Founders FundLux CapitalBain Capital VenturesValor Equity PartnersNvidia
Infrastructure
GPU FleetNVIDIA H100 SXM, NVIDIA H100 NVL, NVIDIA A100 80GB, NVIDIA L40S
Network FabricNVIDIA Quantum-2 InfiniBand
ConnectivityNot specified
StorageNVMe, Ephemeral Disks, Persistent Disks, Shared Disks
Data Center TierProprietary modular data centers co-located at energy sources
Bare MetalYes
AvailabilityOperational
EnterpriseResearchAI/ML StartupsGovernment
Compute & Deployment
On-DemandYes
Spot / InterruptibleNo
Reserved InstancesYes (1-year and multi-year commitments available for enterprise customers)
Bare MetalYes
VM-BasedYes
Container-BasedYes (Docker)
KubernetesYes (managed K8s via Crusoe Kubernetes Service)
Serverless GPUNo
Spin-Up Time2-5 minutes
TerraformYes (official provider)
GPU Hardware
Latest GenH100 SXM, H200 SXM
Legacy SupportA100 SXM
Multi-GPU NodesYes (up to 8x per node)
Max GPUs/Node8
NVLinkYes (NVLink on SXM nodes)
InfiniBandYes (HDR/NDR InfiniBand)
PCIe vs SXMSXM only
HGX PlatformYes (HGX H100 8-GPU)
Pricing Model
Per HourYes (primary billing unit)
Per MinuteNo
SubscriptionYes (reserved instances with term commitments)
Reserved DiscountUp to 45% off with 1-year commitment
Spot DiscountNo spot pricing
Public PricingPartial (some GPUs listed)
Hidden FeesNone disclosed
Pay-as-you-goYes
Credit SystemNo
Performance & Scaling
Multi-Node TrainingYes (multi-node clusters with NCCL support)
Elastic ScalingManual only
Auto ScalingNo
InfiniBandYes (HDR InfiniBand on H100 SXM5 clusters)
NVSwitchYes (on SXM nodes)
Perf IsolationYes (dedicated bare metal)
Noisy NeighborYes (bare metal, no sharing)
Developer Experience
OnboardingSelf-service sign-up with web console; deploy in minutes for standard instances
FrameworksNot specified
SDK LanguagesPython
CLI ToolingBasic CLI for instance and resource management
JupyterVia SSH port forwarding or pre-configured environments
TemplatesPyTorch Training, LLM Fine-tuning, Inference
DocumentationComprehensive docs with API reference, guides, and tutorials
API FeaturesPrometheus-compatible query API
Security & Compliance
SecurityNot specified
ComplianceNot specified
Backed by NvidiaSeries D funded ($686M+)Partnerships with major AI labsFocus on sustainable/green computingUsed by leading AI companies including Mistral AI
Data Center Locations
Coverage
CountriesUnited States
CitiesCorsicana TX, Cheyenne WY, Ellensburg WA, Monclova (Mexico - near US border operations)
North America
Compliance Regions
EU Data ResidencyNo EU presence
US Gov CloudNo
India RegionNo
Datacenter Locations
Key Strengths
Sustainable energy model using stranded natural gas reduces carbon footprint
Competitive H100 pricing due to low energy costs
Purpose-built AI cloud infrastructure
Strong focus on large-scale training clusters
Backed by Nvidia and top-tier VCs
Known Limitations
Primarily US-based regions; limited international availability
Smaller ecosystem and tooling compared to hyperscalers
No spot/preemptible instance pricing
Less mature managed ML services compared to AWS/GCP/Azure
Newer platform with smaller community
Additional Information
Support Options
["Enterprise-grade support","Early access programs support"]
Community
Limited public community presence; primarily enterprise-focused with Slack support channels
Green Energy
Stranded/flare gas utilization and renewable energy focus; marketed as low-carbon GPU cloud
Core Proposition
Crusoe provides GPU cloud compute powered by stranded and flared natural gas, positioning itself as a lower-carbon alternative to traditional data center energy sources for AI workloads.
Notable Customers
Mistral AI
Character.ai
Weights & Biases
Payment Methods
Credit CardWire TransferInvoice (enterprise)
Last updated March 2026. Information subject to change.