I use the assistant in the 4-o version integrated with Typebot to solve doubts about my website, but the context tokens are extremely expensive, and considering they are simple and predetermined answers, context is practically unnecessary. I saw that the API has updated and now allows for limit parameters, but I didn’t understand how to use them, and I don’t know if I can add this directly in the assistant’s settings (if this option doesn’t exist, and from what I searched it indeed doesn’t, it is already a suggestion) or how I should implement the limit directly within Typebot. I will leave images to illustrate.
1 Like