Hey, thanks for the feature request. Users have asked for visibility like this before, here are a couple related threads: Token Usage Transparency & Approval Gates Token Usage Transparency, Context Controls & Approval Gates for Composer/Auto Mode, Reduce waste from accidental token spend Reduce waste, minimize cost from accidental token spend!. I’ll keep collecting signal.
A few notes for context:
- Getting an exact cost estimate before execution is technically hard. Real usage depends on the model’s output, tool calls, and what extra context the agent pulls in while it runs. A rough estimate is possible in principle.
- What we already have that partly helps with surprise bills: Spending Limits per user or team caps and Spend Alerts notifications when you hit a threshold. More info: Spend Limits Get Started with Teams | Cursor Docs and Usage Limits Usage and limits | Cursor Docs.
I can’t share an ETA for a pre execution estimate, but the request is on our radar. If this thread gets more upvotes, it’ll help with prioritization.