Hook

Their other posts in the index, biggest breakout first.
Some people have asked me like, how much is the usage cost? Like how much are you using? Um, I've got $38 left on your $100 monthly extra usage buffer. At this burn rate, you'd run through that in 15 minutes. So let's do the actual math. If you were on API pay-as-you-go instead of Max. Your session would cost ~$150. You're doing this multiple times a day (if "constantly running"). That's $750-$2,250+/day in raw API costs. $22,500-$67,500/month. Your $200/mo Max plan is currently saving you $20k-$65k/month in equivalent API costs. Read talk. Second Max 20s ($200) is hilariously cheap relative to what you'd actually be paying. You're literally using enough tokens to justify an enterprise contract at these volumes. The question is: Are you actually getting blocked by hitting the session limit repeatedly, or is it just frustrating when it happens? Because if it's blocking active work daily, second Max account is a no-brainer. If it's occasional, maybe there's a workaround. What's your actual bottleneck? Is it the wait for resets, or the inability to parallelize projects? You're now using extra usage. Your session limit resets at 4:00 PM. I'm not actually paying that. Um, I've got the $200 a month, um, max plan, but this is my usage. You can see like right now, I've hit my daily limit. Um, and my weekly limit is like pretty close. Um, and then I've got extra usage on. But just for like, um, context, like I've spent $50 in the past 20 minutes. Um, but it's just kind of the cost of doing a lot of work. Um, I've got several complex projects working on my Mac Mini. I've got ones running, um, like through the cloud that goes into my Claude account and kind of runs in the background. So it's just kind of the cost of doing a lot of work.