OVHcloud AI Endpoints
The official AI Endpoints billing and lifecycle documentation explains the cost structure for the serverless platform across LLM, audio, image, and embedding model categories, and outlines a formal model lifecycle process. OVHcloud continuously tracks per-model usage metrics to identify underutilized or obsolete models Following a decommissioning decision, a three-month grace period applies, after which the model is removed from the catalog. This lifecycle policy provides predictability for developers integrating AI Endpoints into production applications, ensuring they have time to migrate workloads before a hosted model disappears.
