I’m using Cursor IDE, and I’m trying to understand the cost of individual AI queries (e.g., chat, code generation, or agent actions).
However, I’ve run into a few issues:
There is no visible “Usage” or “Billing” section in Cursor IDE settings.
I don’t see token counts per request or any per-query cost breakdown in Cursor.
I can see token counts per request or any per-query cost on https://cursor.com/dashboard/usage but I don’t see how to match those counts/costs to my actual LLM requests in Cursor IDE: I can try to guess based on the time, but this is approximate and tedious.
How can I determine the per-query cost of a given LLM request in Cursor IDE?
https://cursor.com/dashboard/usage is interesting but doesn’t easily allow to determine the per-query cost of a given LLM request made in Cursor IDE (we can just to do so trying to match times and guessing prices):
You can’t besides just going to the dashboard right after a request finishes and seeing the latest charge, may be multiple if using subagents (check timestamps). Be lucky you cans still see the actual costs in the dashboard per request. There are reports of even that being hidden.
Hey, good question. This is one of the most common requests on the forum.
Right now the only way to see the cost of a specific request is the dashboard https://cursor.com/dashboard/usage, which you already found. It shows the timestamp, model, tokens, and cost for each request. But you’re right, there’s no direct link between a dashboard entry and a specific chat or request in the IDE. Matching by time is the only workaround for now.
Per-request cost tracking directly in the IDE is a popular feature request, and the team is aware. A couple related threads:
For now, an easier way to track it is to open the dashboard right after an important request. The latest entry should match that request, with the caveat that subagents can create multiple entries for a single request.
