Kubeflow

KServe Inference API

KServe (formerly KFServing) provides a serverless model inference API on Kubernetes, supporting standardized prediction protocols, autoscaling, and multi-framework model serving.

GitHub OpenAPI

Documentation

📖
Documentation
https://kserve.github.io/website/modelserving/v1beta1/

Specifications

⚙
OpenAPI
https://raw.githubusercontent.com/kserve/kserve/master/docs/predict-api/v2/rest_predict_v2.yaml

SDKs

📦
GitHubRepository
https://github.com/kserve/kserve

OpenAPI Specification

rest_predict_v2.yaml Raw ↑