So I paid the monthly but now it asks for more monies for the API accessβ¦..
Discussion
Any suggestions nostr:npub1q3sle0kvfsehgsuexttt3ugjd8xdklxfwwkh559wxckmzddywnws6cd26p ?
Should I purchase the API credits from them or does an unlimited api credit thing actually exist
I would recommend you do not use a predatory subscription plan.
I build Nostr apps frequently with Shakespeare and spend between $2-10 per project. It can become hundreds for highly involved βprojects, but I am not sure what you did to trigger that.
Was it cause it was too complex of a project? Also which model should I be using?
Claude Sonnet 4
Project is not too complex. But it seems it got off to a bad start. It's usually cheaper and more effective to create a new project with a revised prompt if the original is stuck. Or if you're in deep, have it completely remove the feature that isn't working, then add it back.
Thatβs what Iβm using but idk why it sucks up so much $$$ just to debug and not actually debug
https://video.nostr.build/ac1312ec280231684e7d49f735624532cef158ec2f0699aa552e41652124253f.mp4
That one prompt was $17
Once more it still hasnβt removed the mock data and the buttons still not working properly )):
https://video.nostr.build/02b24fd5113d0e640a5de6b408ce58995fefd9ffa0101dd1e853759e8be76ff3.mp4
if you have a way to restart the context, then do that.
typically, you just start a new chat with the robot to accomplish such things.
update: i think it was in demo mode??
https://video.nostr.build/e266f8cdc86d6eec131af54f55ac6685c51f913c89633379b29a77af2224be0d.mp4
lmk if it works for you https://nostr-secret-santa.shakespeare.wtf
nostr:npub1hee433872q2gen90cqh2ypwcq9z7y5ugn23etrd2l2rrwpruss8qwmrsv6
Going through some of your prompts, your app required lightning and privacy
These are can be tricky to vibe code, but are core functions weβre working to improve
nostr:npub1yzfm42rzr3dj2h50flpvdl0uzrv22kv2y4ghve804w5xqu6lzqcqkyfxu5 is working on vibe codeable private messaging and Iβm trying to improve vibe codeable bitcoin
I think the original model you started with (GLM-4.6) got you off to a poor start with that level of prompt complexity and it may have snowballed from there
Weβve since removed that model based on your feedback and will continue to make it better for you π«‘
yo wtf, that sucks lol
these subscription hustlers are getting worse by the day... first they charge monthly, then need credits on top? classic bait n switch
look, if you're burning cash just to debug some mock data, might be time to yeet the whole project and restart with a clean slate. or honestly? hit up the open source crew - could prolly knock this out in a weekend with some proper code instead of burning benjamins on ai that can't even remove a hardcoded array π
Iβll try building a new project from scratch with the correct model and a much less complex structure just to see if we can get something going and we can later go from there! Thank you again Lemon π« appreciate your work!
Of course, homie
Appreciate you trying again
Weβll make it better for sure, itβll be able to do your original prompt soon enough π
Hereβs some resources that may help (not sure if someone already sent this):
Of course, canβt wait bro! π₯
Alex shared it Iβm taking a look at it now π«‘
cc: nostr:npub18ams6ewn5aj2n3wt2qawzglx9mr4nzksxhvrdc4gzrecw7n5tvjqctp424 seeee i did try Shakespeare
Claude code is not predatory at all. Itβs far cheaper than burning thousands on API. Demonizing subscriptions is quite foolish and ivory tower of you lol, with your millions in funding.
Spending $2-$10 on a project = slop pump and dump project that no one is going to use.
Pure LARP
Why the hate π