What’s the difference in the 8k and 32k model of gpt 4?

i’ve gotten access to gpt 4 but i am confused on what i’m using. for the api playground, the model just says gpt 4. is this the 8k model? if not, how do i access and use the 8k model? the limit is still 2000 tokens so i think it’s not the right one. could anybody please explain further what is what?

There may be some confusion right now in the Playground.
Davinci-003’s settings is accepting 4K tokens, meanwhile the GPT-4 chat remains at 2k.

Something interesting to note:

An API call to Davinci-003 with a max token length of 4,000 is accepted. Previously it was only 2,000.

Looking at the docs (which have been separated now), there is a new notification under the token length parameter: ( keep in mind, this is the completions docs, not the chat completions )

The token count of your prompt plus max_tokens cannot exceed the model’s context length. Most models have a context length of 2048 tokens (except for the newest models, which support 4096).

To answer your question, as of now it seems that the only way to get the 8k token version is by your own API request.

Thank you for this question. It makes me believe that Davinci is currently being worked on.

thank you for your reply. I guess now I am forced to stop procrastinating and work on my programs lol