$200 Grok Bot plan burns 99% of weekly quota in three days
An Ultra user reported using 65% of the weekly allowance on day one, 84% on day two and 99% on day three, with about 84% of tokens going to cache re-reads; the author suggests a fresh chat per task and fewer scheduled routines.
Holy sh*t, I found a f@cking number that explains why a $200 Grok Bot plan hit 99% in 3 days A $200 GROK BOT PLAN HIT 99% OF ITS WEEK IN 3 DAYS. 84% OF HIS TOKENS WENT TO ONE THING xAI's page for Grok Bot puts the pitch in two lines: "Ask a Bot to follow along as you complete a workflow once. It saves it as a routine and runs it on its own next time." It comes back only when it needs your approval. A user on the $200 Ultra plan posted his meter on the Cursor forum: 65% of the weekly allowance on day one, 84% on day two, 99% on day three. His token export for days two and three showed about 284 million tokens, roughly 84% of them cache reads. That's the bot re-reading history it already has. My math: about 239 million tokens spent re-reading, not working. Long chats make every later run more expensive. Schedules add to it. Every routine run costs usage, even when it finds nothing. One every 5 minutes fires 288 times a day. Hourly is 24, twelve times fewer. I think the first fix is the chat, not the plan: one fresh chat per task, and routines on the clock you actually need. Watch the meter go 65, 84, 99, with the run counter behind it.
This post quotes another post: x.com/kocer_eth/status/2092216029816472002