There was no included warning that this would be a hard switch and now the entire project I was working with is stuck on the 3.5 engine which effectively will make me have to redo a weeks worth of activity to get the 4.0 engine back up to speed. I haven’t capped before and didn’t expect a sudden cut off with no warning nor did I not expect to be able to resume the 4.0 engine. Is there anything in the pipeline to get this fixed? Or at least a warning that the cap is approaching? Upgraded plan past the Plus? A heavy warning saying the conversation will be permanently switched to the 3.5 engine? For any kind of coding or deeper technical research really suffers from this kind of thing. Funny that I never realized I was paying for a capped service as this was changed after I was let into the service and never really payed attention like I am now.
I have had this happen before too.I’ve not heard of anything to address it. I’ve definitely learned my lesson not to switch though. I’m sure it’s an oversight that will be patched at some point.
Do you have api access by chance? If you apply for api access to GPT4 you can get past the cap and control your conversation better with a custom UI but then you are paying for access beyond your normal subscription.
Excellent point - I didn’t think to use the API, I figured it’d be on the same metering style.
Guess I know what I’m doing tonight - thanks for that.
No prob! Happy coding. There is nice working ui and lite version by McKay. There is also a library somewhere that has some chat stuff in it. I’ll see if I can get you a couple links.
While I get nobody wants to hear this, but there is a very low hanging and trivial solution:
Raise the prices.
It’s true this will increase demand for alternative solutions, but the complete lack of availability of the API is likely raising demand for alternative solutions even more.
Maybe make the prices based on supply / demand. It can be manually tweaked once a month - don’t have to automate it.
Does the 25 messages per hour cap gets bypassed with api?
I’ve never been limited with the API when just using GT4 as part of my workflow.