I have performed an API call to OpenAI's endpoint https://api.openai.com/v1/models .
The endpoint lists the currently available engines, and provides basic information about each one such as the owner and availability.
As a logged-in user, I get a JSON response of 63 models. These are the most recent ones (currently) , formatted, shown with release date.
59: "11/28/2022, 2:40:35 AM : text-davinci-003"
60: "12/16/2022, 8:01:39 PM : text-embedding-ada-002"
61: "2/27/2023, 10:13:04 PM : whisper-1"
62: "2/28/2023, 7:56:42 PM : gpt-3.5-turbo"
63: "3/1/2023, 6:52:43 AM : gpt-3.5-turbo-0301"
I notice that there are 2 very similar models , "gpt-3.5-turbo" and "gpt-3.5-turbo-0301", with gpt-3.5-turbo-0301
released only 11 hours after gpt-3.5-turbo
.
What is the difference between these two model versions? It does not seem to be a glitch or a misnaming error. Why did OpenAI bother to include both of them, and why didn't take the inferior version?
(I haven't experimented with these two models in any way yet. I might do this very soon. However I though I might as well ask here. Informing others in this forum might have some benefit.)