Hi everyone,
does anyone know if there’s a clear table or page in Cursor that lists the exact prices for each model?
I understand that some models have, for example, 4× request cost, others 2×, depending on the settings (e.g. when not using “max”), and some might have different multipliers.
Recently, I started using grok-fast-code-1, which significantly helps reduce costs when I’m working on simple tasks and give very specific implementation instructions. However, I’d really like to understand how much models like Grok-4, GPT-5, cloud-sonnet-5-thinking, or max actually cost in practice.
For example:
-
Is GPT-5 always charged at $0.04 per request?
-
Is Grok really $0, or is there some hidden multiplier depending on usage?
-
How does cloud-sonnet-5-thinking compare in pricing?
-
If I use 140 million tokens with Grok-4, how much exactly would that cost compared to GPT-5 or cloud-sonnet-5-thinking?
It feels like this pricing information changes so often that it’s basically impossible to keep track. And I don’t think the right approach should be to just “try the model and see what you’re charged afterwards” — normally, you know upfront what you’re paying for.
So my main question is:
- Is there any official, up-to-date table or page on Cursor’s website where I can clearly see the token pricing and multipliers for each model, including Grok, GPT, Sonnet, and the various “max” versions?
Right now, this info is quite hard to find, and since there are so many models with different behaviors and multipliers, it’s very difficult to plan costs in any predictable way.
I’m not trying to complain — I just want a transparent, centralized source of pricing information for all models so I can plan properly.
Thanks a lot for any help or links!