@therubberduckiee: Thought about this the other day, thought it was interesting and helped me distuinguish how we are actually getting charged on requests, figured it would be helpful to make a video on it. #ai #pricing #tokens #prompts #code
This is true but needs to be used carefully in the context of Agentic AI. Max tokens may truncate the output of the intermediate sub-prompts from the agent leading to context loss.
2025-09-08 17:54:01
2
🆘🇺🇸🏳️⚧️🏳️🌈🇵🇸 :
Great question, based on how cursor works I believe you are correct.
2025-09-08 18:17:53
0
andyccs :
Just prompt it to summarize it in 1000 token
2025-09-09 20:13:39
0
obamna :
warp mentioned 🗣️
2025-09-09 02:35:40
0
Luke Rhodes :
Absolutely the 10k but asking it to think more about how to be efficient with its output including fewer comments, while maintaining best practices and code clarity can definitely save some cash. No way I’d be paying for the Opus API though - Claude Max if you can justify the cost
2025-09-18 08:47:56
0
Shai-Shalom Hadad :
Rule of thumb: it’s never the cheaper answer.
2025-09-09 08:23:19
0
princeomar :
So with that rule will your output be less than 1000
2025-09-09 00:02:35
0
The Synthetic Mind :
Yes, you are right.
2025-09-08 17:33:10
0
Automated Intelligence :
Correct
2025-09-08 23:32:29
0
Jim :
Dumb question, but is this limit easy to configure in either cursor or Claude code?
2025-09-09 02:08:47
0
santiago18150 :
Yes and water is wet
2025-09-08 21:43:45
1
To see more videos from user @therubberduckiee, please go to the Tikwm
homepage.