2 comments

[ 3.3 ms ] story [ 19.1 ms ] thread
Does this use the 32K token context model of GPT-4 on the back end?
Not yet, but we will migrate to gpt-4 soon