Nebius

Inference: run chat, completions and embeddings on models, plus datasets, fine-tuning and dedicated endpoints

Deployments: list NIM services, look up their GPU metrics, and check GPU and service health