I ran out of fast model requests. I still have gpt-4o-mini or cursor-small requests left.
I can use these requests in the inline editor, or chat, by selected the models. However, autocomplete is no longer working.
I tried going to Cursor Settings>Models, and disabling everything but gpt-4o-mini and cursor-small. However, that didn’t work in terms of making autocomplete available again.
Is there any way to have autocomplete with the non-premium models?
Hi @SolanK
Autocomplete uses a custom Cursor model that’s different that gpt/claude.
Are you using the free plan?
Yes, I’m using the free plan at the moment.
With the free plan, you only get access to 2000 completions.
Please note that autocomplete uses a custom Cursor model that’s different that gpt/claude.
OK, I upgraded to Pro and autocomplete worked. Thank you.
Is there a way to see the number of completions you’ve used in the cursor.com Settings? Under “Usage” it shows the number of “Premium models” and “gpt-4o-mini or cursor-small”, but I don’t see completions.
You don’t need to worry about it, they’re unlimited on the pro plan.
OK thank you 
But out of curiosity, is there any way to see completions use? Or is that hidden?
As far as I know, this is hidden, but you can check some logs here, though they probably won’t be of much help to you.
Great thanks a ton for your help
Is cursor-small now finetuned on gpt-4o-mini?
