API reference
Dedicated endpoints
17 operations. Each page has the request, the response, the errors and examples in curl, TypeScript, Python and the CLI.
- GETList dedicated endpoints in the project, newest first
/v1/endpoints - POSTCreate a dedicated endpoint (a catalog model or your own Hugging Face model) with its first deployment
/v1/endpoints - GETGet a dedicated endpoint with its deployments and replica counts
/v1/endpoints/{dedicated_endpoint_id} - PATCHUpdate the display name or the inactivity timeout
/v1/endpoints/{dedicated_endpoint_id} - DELETEDelete an endpoint: its replicas are removed and its model ID stops resolving
/v1/endpoints/{dedicated_endpoint_id} - POSTStart a stopped endpoint (needs credit; replicas start and are billed once ready)
/v1/endpoints/{dedicated_endpoint_id}/start - POSTStop an endpoint: its replicas are removed and billing for them ends
/v1/endpoints/{dedicated_endpoint_id}/stop - GETAn endpoint's deployments
/v1/endpoints/{dedicated_endpoint_id}/deployments - POSTAdd a deployment (other weights, engine or hardware). It starts with no traffic; give it a share with the traffic split.
/v1/endpoints/{dedicated_endpoint_id}/deployments - GETGet a deployment
/v1/endpoint-deployments/{deployment_id} - PATCHChange replica bounds or the autoscaling target (replicas are not recreated)
/v1/endpoint-deployments/{deployment_id} - DELETEDelete a deployment that carries no traffic; its replicas are removed
/v1/endpoint-deployments/{deployment_id} - PUTSet the share of requests each deployment receives (weights sum to 100)
/v1/endpoints/{dedicated_endpoint_id}/traffic - GETAn endpoint's replicas, newest first
/v1/endpoints/{dedicated_endpoint_id}/replicas - GETWhat happened to an endpoint, newest first
/v1/endpoints/{dedicated_endpoint_id}/events - GETRequests, errors, tokens, latency and billed replica time, by deployment and over time
/v1/endpoints/{dedicated_endpoint_id}/analytics - GETPublic catalog models a dedicated endpoint can serve, with the engine used on CPU and on GPU hardware
/v1/endpoint-models