KEDA / GUIDES
What is a ChatGPT API relay? A practical introduction
Understand API relays, OpenAI-compatible endpoints, web chat and ChatGPT subscriptions, and what to check before choosing a service.
How an API relay works
An AI API relay receives requests from an application, forwards them to an upstream model according to its configuration, and returns the response. OpenAI compatibility describes a request format; it does not imply OpenAI ownership or identical features across models.
Web chat, APIs and subscriptions
Web chat is useful for direct questions and creative work. APIs connect applications and automation. ChatGPT subscriptions and third-party API credit are managed separately. Keda brings aggregated chat and developer access into one platform; the console lists your permissions, available models and charges.
What to check before connecting
Check model IDs, supported endpoints, input and output pricing, data policies and error responses. Start with a small amount of non-sensitive test data and verify both regular and streaming requests. Keep API keys in server-side environment variables, away from browser code, public repositories and screenshots.