CloudGPU

Name: CloudGPU GPU Cloud
Brand: CloudGPU
Availability: InStock
Rating: 4.8 (156 reviews)

⚡ GPU Rental Available Now

CloudGPU is a GPU cloud for developers and the AI agents they run. It offers 19 GPU models, including RTX…

🏢 Taiyuan, China📅 Since 2025🌐 Website ↗

Min Rental

Hourly (no min)

Setup Time

Spot Instances

✓ Available

Support

Email Support

CloudGPU is a GPU cloud for developers and the AI agents they run. It offers 19 GPU models, including RTX 5090, RTX 4090, A100, A800, H800, H20 96G and Ascend 910B. Prices range from $0.15 to $1.79 per hour across Asia-Pacific. Every instance comes with root SSH and local NVMe/SSD storage. You can launch from ready templates for Ollama, ComfyUI, JupyterLab or a raw GPU. CloudGPU charges by the minute, with a one-minute minimum. The meter stops when you destroy the instance. One prepaid balance covers everything, and you can pay by card or USDT. The same account gives you an OpenAI-compatible model API. A REST GPU API and an MCP server let your agents rent and manage GPUs. CloudGPU does not offer persistent storage or spot instances yet. Live GPU prices: https://cloudgpu.app/gpu-prices.json?utm_source=computestacker&utm_medium=directory&utm_campaign=gpu_pricing . Pricing page: https://cloudgpu.app/pricing?utm_source=computestacker&utm_medium=directory&utm_campaign=gpu_pricing

Pros & Cons

Pros

Per-minute billing, rounded up
Root SSH
Templates for Ollama, ComfyUI, JupyterLab and raw GPU
OpenAI-compatible model API at cloudgpu.app/v1
MCP server (npx -y cloudgpu-mcp) and a REST GPU API
Payment by card or USDT

Cons

No persistent storage
No spot instances
No SLA

Pricing Data Not Scraped

Exact pricing for this provider must be viewed directly on their website or by requesting a custom quote.

💡

Pro Tip: Spot vs. On-Demand Pricing

Rates shown above are standard on-demand pricing. This provider offers spot instances, which can reduce your costs by 30-70% if your workload is interruptible. Request a custom enterprise quote →

Core Infrastructure

Available GPUs

RTX 2080, RTX 3090, RTX 4090, RTX 4090D, RTX 4090 48G, RTX 5090, RTX 6000D, RTX PRO 6000 96G, RTX A6000, V100 32G, A30, A40, A100 40G, A100 80G, A800 40G, A800 80G, H20 96G, H800, Ascend 910B

Network Interconnect

NVLink

Storage Options

local disk with each instance

Data Center Locations

Asia-Pacific