Engine Names & Documentation Update

Hello! The Instruct series, previously in beta, are now our recommended GPT-3 models.

As part of this change, we’ve also updated many of our engine names to be clearer and support adding more models in the future. API requests that specify the old engine names will continue working as expected, but we recommend that you start using the new names going forward.

The changes you will see:

  • Engine names will no longer include the “instruct” qualifier, and instead use our new naming convention (reflected in this help article)
  • Documentation will reference the Instruct models
  • The playground will default to Instruct models
    • Search, fine-tuning, and embeddings should still be used with our base GPT-3 models (davinci, curie, babbage, and ada)

As always, don’t hesitate to reach out if you have any questions or feedback!


I was wondering if there is, or will be, a way to rename our fine-tuned engines? At the moment I see them come through as something like “ada:ft-personal-” which isn’t too clear what the enginehas been trained on.