One network for inference and GPU compute. Pick where to start; the full dashboard is a click away.
Call open models on one OpenAI-compatible endpoint, paid per token, with a key and a base URL.
Launch a VM with whole GPUs passed through, by the hour. SSH in and run whatever you like.
Apply to become a provider. Once approved, one command enrolls a box, and you earn on the paid hours it hosts.
NVIDIA or AMD GPUs with an outbound internet connection.