whats the difference in max output between ChatGPT Window or using the API chatgpt4o…i can only see the api 4,096 tokens …but it seems to not state anywhere what openai has limited chatgpt 4o window too
2 Likes
Yeah, there’s a bit of “black box” stuff for ChatGPT, I’m sure.
The new o1 model seems to output more content easily…
As from october that same table says the GPT 4o should use 160xx tokens as max output limit. Also the term window is confusing, as the column header has “context windows” and the corresponding value is 128000 tokens.
The post about the “backbox” composition of another backbox might be right on. The web interface obscurity seems to allow selling one thing on the hope that it could be throttled arbitrarily for some just in time logistics, or something I can’t fathom, but seems to be very important so to make the bots more absurd than they would be if left alone.
I don’t think it might be respecting the intelligence of the users that way.
but that is the prerogative of such business model.