PSA: Turning off my model's default system prompt cut my token bill by 41%

I was burning $60 a month on API calls for a small chat app until I stripped the 800-token default system prompt down to two lines, and the replies actually got tighter. Anyone else find that less prompt bloat beat every fancy framework you tried?

1 comments

Log in to join the discussion

Log In

1 Comment

grant_dixon6d ago

800 tokens of "you are a helpful assistant" fluff is wild. Tbh I ran the same test on my side project and the replies got shorter and clearer once the default prompt was gone. My bill dropped from around $40 to $25 that month, mostly because the input side shrank. Less filler in, less filler out.

2