2 comments

[ 2.1 ms ] story [ 18.5 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