This article is only relevant if you started using the API before June 6, 2022.
We are deprecating the term ‘engine’ in favor of ‘model’. Most people already use these terms interchangeably, and we consistently hear that ‘model’ is more intuitive.
Moving forward, API requests will work by referencing a ‘model’ instead of an ‘engine’. If you have used a fine-tuned model, then you are already familiar with using ‘model’ instead of ‘engine’ when making an API request. Engine listing is also being replaced by Model listing, which will consolidate both base and fine-tuned models in a single place.
We will maintain backward compatibility for requests using ‘engine’ as a parameter, but recommend updating your implementation as soon as you can to prevent future confusion.
For example, a request to the completions endpoint would now be (full details in our API reference):
Deprecated | Current |
response = openai.Completion.create( | response = openai.Completion.create( |
openai api completions.create -e text-davinci-002 -p "Say hello world three times."
| openai api completions.create -m text-davinci-002 -p "Say hello world three times."
|
curl https://api.openai.com/v1/engines/text-davinci-002/completions \ | curl https://api.openai.com/v1/completions \ |
We have updated endpoint URL paths accordingly (full details in our API reference):
Deprecated | Current |
https://api.openai.com/v1/engines/{engine_id}/completions | https://api.openai.com/v1/completions |
https://api.openai.com/v1/engines/{engine_id}/embeddings | https://api.openai.com/v1/embeddings |
https://api.openai.com/v1/engines | https://api.openai.com/v1/models |
https://api.openai.com/v1/engines/{engine_id}/edits | https://api.openai.com/v1/edits |