RunPod
GPU cloud for inference and fine-tuning — on-demand pods and serverless endpoints that scale to zero.
Overview
RunPod rents GPUs two ways: pods, which are persistent machines for training and experiments, and serverless endpoints, which spin a container up per request and down when idle. Bring a Docker image, pick the GPU class, and the endpoint handles queuing and autoscaling. Commonly used to self-host open-weight models behind an OpenAI-compatible API when the hosted providers are too expensive or the data cannot leave your control.
Why it's indexed
The usual first stop for teams moving an open-weight model from a laptop to something an agent can call in production.
Editorial judgement, not a measurement — see our methodology. Listing is never paid for.
At a glance
- Vendor
- RunPod
- Category
- Hosting & Deployment
- License
- Proprietary
- Deployment
- Hosted
- Languages
- Python, Any
- Pricing model
- Usage-based
Pricing is shown as a model, not a rate. Published rates change without notice — check the vendor's own page before budgeting.