RunPod
RunPod is an AI-focused cloud infrastructure provider offering GPU compute for machine learning workloads. The company specializes in on-demand GPU instances, serverless AI workloads, and multi-node GPU clusters across 31 global regions.
AI
88/100
From $0.38/hr
Fresh ai-ml gpu-compute developers enterprise
Rating Breakdown
Performance 5.0
Value 6.9
Features 5.0
Support 6.0
AI Tools 8.8
Hosting Plans
RTX 4090 Pod
gpu $0.38/hr
41 GB RAM storage unlimited bandwidth 1 site
Serverless RTX 4090
gpu $1.10/hr
24 GB VRAM storage unlimited bandwidth 1 site
A100 SXM Pod
gpu $1.79/hr
125 GB RAM storage unlimited bandwidth 1 site
H100 PCIe Pod
gpu $1.89/hr
188 GB RAM storage unlimited bandwidth 1 site
Serverless H100
gpu $4.18/hr
80 GB VRAM storage unlimited bandwidth 1 site
H200 Pod
gpu $4.31/hr
276 GB RAM storage unlimited bandwidth 1 site
B200 Pod
gpu $7.56/hr
283 GB RAM storage unlimited bandwidth 1 site
Infrastructure
Control Panel custom
Uptime SLA 99.9%
SSH Access Yes
CDN Included No
Git Deploy Yes
Datacenters 31 global regions, 8+ worldwide regions
Support
Channels email, ticket, live-chat
24/7 Support No
Software Support
Languages Python, Docker, PyTorch, TensorFlow, CUDA
Pros
- Specialized GPU infrastructure with H200, B200, H100, and other high-end GPUs
- Sub-200ms cold-start times with FlashBoot technology
- Serverless scaling from 0 to 1000+ workers in seconds
- No ingress/egress fees on S3-compatible storage
- Competitive pricing with per-second billing
Cons
- Primarily focused on AI/ML workloads, not general web hosting
- Requires technical expertise in machine learning and GPU computing
- Limited support for traditional web hosting features
- Pricing can be complex with multiple GPU options