Does Fooocus have a hard or soft token limit ? I have noticed that quality drops off the longer the prompt is.

[swamp-cabbage]

I'm very impressed with the output of Fooocus, I use it in conjunction with SD.Next to generate "Influencer" kind of content (realistic human beings).

When prompting in SD.Next (or A1111) there is a token count at the top right of the prompt input boxes, so you can see how many tokens you are working with.

I have noticed that in Fooocus, short prompts are really better and more effective (by design, I assume).

My question - is there a limit to prompt tokens that Fooocus takes into account ?

Thanks !