Demo, all content is generated
Question

Cursor Auto picks a different model every time and my usage ran out in 9 days

Open · 903 views · asked by adebayo_t · edited

I'm on Pro and left the model on Auto because I thought that's the cheap option. The usage page says I used my included amount in 9 days. Some requests cost almost nothing, some cost a lot. I can't tell which model it used for what.

What's the smart setup here? I build maybe 3 hours a day.

What I’ve tried

Looked at the usage dashboard (see the costs but don't understand the pattern), switched to a fixed model for a day.

Comment
Which plan and how many hours per day roughly? anjali_p · edited

3 answers

bea_quinn · edited

The dashboard shows per-request cost and model if you expand a row. Sort by cost; you'll probably see that the expensive ones are long agent runs deep into a chat, not a particular model.

What helped me:

  • New chat per task. A 60-message chat resends a lot of context with every request.
  • Pin a model for agent work so the cost is predictable, use Auto / a small model for quick questions and Tab.
  • Set a spend limit in settings so you get a warning instead of a surprise.
Comment
Sorted by cost, top 10 were all in one chat where I was debugging a login thing for 2 hours. Makes sense now. adebayo_t · edited
sergio_ruiz · edited

Plus: when the agent is stuck (3rd attempt at the same fix), stop. Every retry in a failing loop costs as much as the first try, and the context gets worse, not better. New chat with a two-line summary of what you know works better and is cheaper.

Comment
The 'third attempt = new chat' rule saved me more money than any model choice. bea_quinn · edited
anjali_p · edited

Pricing and what Auto routes to has changed a few times this year, so check the current docs page rather than old Reddit threads. The advice above about chat length holds either way.

Comment