Skip to main content
Version: 19.x (unreleased)

Inference Secret Resource Reference

Report an issue with this page

The inference secret resource configures a secret, such as OpenAI API key or LiteLLM master key, that is used along with an inference_model. Multiple inference models can use the same secret. Inference secret resources can be listed and retrieved, but their values will be stripped from the payload to protect them from leaking outside.

kind: inference_secret
version: v1
metadata:
  name: example-openai-key
spec:
  # value is the secret value, e.g. an OpenAI key.
  value: "************************************"