Hi, sometimes auto takes tokens (and credits, see screenshot). I’m a pro-user. I already tried to argue with the email support ai that it shouldn’t or at least get a better understanding of how auto works. When it couldn’t, it refered me to a human that hasn’t yet responded (which I get, probably loads of questions).
I would be super grateful, if anyone is deeper in the topic and/or could point me to some resources. Thank you!
a Newbie
refering this:
Hi, thanks for the question!
This is unfortunately just a rendering issue in cursor.com which I have passed to the team, but I can assure you the use of ‘Auto’ is unlimited on every paid plan.
Hi Dan,
Is auto still unlimited? Because I noticed even the wording on the dashboard a few days ago specifically removed the line about “Auto usage does not count towards your usage limits.” The auto API cost I see in the dashboard, is that just there for me to know how much the value of auto is, or is it there because it’s part of my monthly usage limits. It appears to be a bit confusing, since it even gets added into the total api cost at the bottom of the chart, so if someone wanted to know what their real usage was, they would have to subtract the auto api cost from this total.
Thanks for the clarification.
Auto is still unlimited afaik. I have the 20$ plan and they let me use 50$ of credit for premium models before telling me I need to auto. Combined with auto I have over 120$ usage in the dashboard and auto continues to work every time (if you can call it “working” whenever you get gpt-4 in auto mode).
Pro tip re premium models: Claude 4 sonnet is currently buggy and using way too many tokens for cache reading. Switch to gemini 2.5 for the time being.
Thanks, that’s reassuring. Don’t know why they don’t make this more clear, it is a great selling point. I am generally happy with how auto works and if I can truly use it unlimited at it’s current state, it’s a great value.
