SubModel GPU Cloud
A Peer-to-Peer Compute Cloud Platform
Globally Distributed GPU Cloud for Your AI Workloads
Headquartered in Singapore with strategic hubs in Hong Kong and New York, SubModel delivers enterprise-grade AI infrastructure through our proprietary global computing network.
Our founding team combines decades of AI infrastructure expertise from Meta's ML systems, IBM's quantum initiatives, NVIDIA's GPU architectures, and AWS's cloud ecosystems.
Global Compute Fabric
  • Proprietary intelligent orchestration engine
  • 500+ optimized GPU clusters across 10+ regions
Unmatched Price-Performance
  • Full-stack support for mainstream GPUs
  • 20% cost advantage vs. RunPod
Next-Gen AI Scaling
  • Trillion-parameter model training infrastructure
  • Enterprise-grade inference and hybrid edge-cloud deployment solutions
Service Type
Maximize your AI budget with our purpose-built GPU cloud infrastructure optimized for performance-per-dollar across workloads.

Pod（On-Demand）

View more

Serverless

View more

Bare-Metal

View more

GPU Pricing

RTX4090

Supply Quantity: 3,000 units Pod (On-Demand) Pricing: $0.30/hour Serverless Pricing: $0.00020/second Configuration: CPU Model & Cores: Intel Xeon 8175M @ 2.5GHz, 112 threads GPU Model: 8 × NVIDIA RTX 4090-24GB Memory Capacity: 512GB Storage Type & Capacity: 2TB + 6TB SSD Network Card: Intel® Ethernet Controller XL710 for 40GbE QSFP+

RTX3090

Supply Quantity: 2,000 units Pod (On-Demand) Pricing: $0.18/hour Serverless Pricing: $0.00015/second Configuration: CPU Model: Intel 6330 @ 2.0GHz 2 GPU Model: 8 × NVIDIA RTX 3090-24GB Memory: 512G Storage: 480GB SATA SSD 7.68TB U.2 NVME SSD Network Card: 25G Dual Port

H100

Supply Quantity: 256 units Pod (On-Demand) Pricing: $2.03/hour Serverless Pricing: $0.00093/second Configuration: CPU: Intel 8480+ (dual CPUs @ 2.0GHz each) GPU: 8 × NVIDIA H100-80GB Memory: 2TB (64GB × 32 modules) Storage: 960GB SATA SSD × 2 7.68TB U.2 NVMe SSD × 8 Network Card: 400G × 8 ports

User Category
Standard User
  • Campaign Incentives
  • 1% Reward Credits
Additional Benefits:
  • Standard Developer (Estimated demand: 1-80 GPUs)
  • Deployment: Pod/Serverless
  • Reward Credits: 1% of last month's total spend
  • Value-Added Service: Community-driven technical support team access
Core User⭐ 22% Off
  • Up to 22% discount
  • 3% Reward Credits
️️️Additional Benefits:
  • SME AI Company (Estimated demand: 80-400 GPUs)
  • Deployment: Pod / Serverless / Bare-Metal
  • Reward Credits: 3% of last month's total spend
  • Test Device: 1 unit (7-day free trial)
  • Value-Added Service:
  • 1v1 dedicated customer service
  • 12/5 technical support
Strategic Partner⭐⭐ 53% Off
  • Up to 53% discount
  • 5% Reward Credits
Additional Benefits:
  • Enterprise AI Leader (Estimated demand: 400+ GPUs)
  • Deployment: Pod / Serverless / Bare-Metal / On-premise Deployment
  • Reward Credits: 5% of last month's total spend
  • Test Device: 3 unit (7-day free trial)
  • Value-Added Service:
  • 1v1 dedicated account manager
  • 24/7 technical Support
  • Critical issue response within 24 hours
Price Advantage
Serverless Pricing Comparison ($/s)
*SubModel pricing is based on the highest discount
Pod (On-Demand) Pricing Comparison ($/hr)
*SubModel pricing is based on the highest discount
Taking the RTX 4090 as an example, ordering 50 units (8 cards/unit) in SubModel can save up to 80% in costs compared to RunPod.*
*Expand to view calculation details
Serverless Delivery Cost Structure
Pricing Formula: Unit Price × Seconds × GPU Count
RunPod: daily cost $ 178.56
SubModel：
  • Standard: daily cost $115.2, equals to 64.5% of RunPod (35% Off)
  • Core: 20% base + 3% rebate at a daily cost of $89.40, which is only 50.06% of RunPod (50% Off)
  • Strategic: 50% base + 5% rebate at a daily cost of $54.72, which is only 30.65% of RunPod (70% Off)
Pod (On-Demand) Delivery Cost Structure
Pricing Formula: Unit Price × Hours × GPU Count
RunPod：
  • Secure Cloud: daily cost $ 6,624
  • Community Cloud: daily cost $ 3,264
SubModel：
  • Standard: daily cost $2,880, equals to 43.5% of RunPod Secure Cloud (57% Off) and 88.2% of RunPod Community Cloud (12% Off)
  • Core: 20% base + 3% rebate at a daily cost of $2234.88, which is only 33.74% of RunPod Secure Cloud (66% Off) and 68.47% of RunPod Community Cloud (32% Off)
  • Strategic: 50% base + 5% rebate at a daily cost of $1368, which is only 20.65% of RunPod Secure Cloud (80% Off) and 41.91% of RunPod Community Cloud (60% Off)
Technical Support
Exclusive VIP Support
24/7 assistance via a dedicated VIP community and private maintenance team.
Customized Solutions
Tailored design, development, and optimization services exclusively for VIP clients.
Private Deployment
Isolated availability zones with dedicated egress lines and bandwidth for exclusive resource allocation.
High-Efficiency Support
Priority escalation, elevated SLA guarantees, and 24/7 maintenance for mission-critical operations.
Technical Highlights

1

High Compatibility
Compatible with standard mirrors, migration requires no rearrangement, allowing you to start immediately.

2

Ultra-Fast Orchestration
Effortlessly schedule global computing power in seconds—no waiting required.

3

Serverless Architecture
Automatic scaling to meet the needs of the entire business cycle.

4

Developer-Centric Tools
Provides API, SDK, CLI, fast access services, and includes reference for open source projects.

5

Military-Grade Security
Implement high-standard security and encryption measures to safeguard customer data.

6

Self-Healing Infrastructure
T3+ level data center, multiple solutions, and automatic repair of single point of failure.
Partners
Made with Gamma