GPU for rent: high-performance GPU marketplace
Every provider's price, side by side.
Compare live hourly prices for cloud GPUs across providers and regions, filter by VRAM, and deploy the one you choose from the OneInfer Console.
Find the NVIDIA H100, A100, RTX 4090 and more, launch it with your own Docker image, connect over SSH, and pay from one wallet while it runs.
Deploy a GPU in 4 steps
Search by GPU name, then filter by provider, region and minimum VRAM, and sort by price or memory.
Compute marketplace
64 of 64 GPUs
Typos are fine (“nvidai”, “hoper”), and * or ? stand in for any characters. Model numbers match as typed.
Explore every GPU model
Compare VRAM, starting hourly prices, providers, and available regions across 63 GPU models.
GPU performance charts
GeekBench score comparison
GPU performance comparison
Training performance comparison
Inference performance comparison
About the GPU marketplace
Whether you are training large language models, running high-throughput API inference, or executing massive data simulations, the OneInfer GPU Marketplace gives you on-demand access to cloud GPUs for rent.
OneInfer connects AI researchers and developers directly with datacenter providers and lists their GPU rental options side by side. Filter by provider, region, and minimum VRAM, and compare specifications such as RAM, PCIe version, and compute generation (such as Hopper or Ampere architecture).
The inventory includes GPUs such as the NVIDIA H100, A100, RTX 4090, and RTX 3090. Launch pre-configured Docker environments, monitor your instances, and only pay for the compute time you consume.