Slow requests even if I have usage-based plan

Bug Reports
1

Describe the Bug

Soo, everything was normal until now. I have usage-based pricing on my account cuz I run out of 500 premium requests and I have slow requests!

Steps to Reproduce

Make usage based requests always fast.

Screenshots / Screen Recordings

Zrzut ekranu 2025-06-15 o 09.45.54.png
Zrzut ekranu 2025-06-15 o 09.45.54.png1155×339 28 KB

Operating System

MacOS

Current Cursor Version (Menu → About Cursor → Copy)

Version: 1.1.2 (Universal)
VSCode Version: 1.96.2
Commit: 87ea1604be1f602f173c5fb67582e647fcef6c40
Date: 2025-06-13T00:26:52.696Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.5.0

Does this stop you from using Cursor

No - Cursor works, but with this issue

1 Like
2

Yeah, I got the same issue. And the limit is not reached for me, and changing it to even higher number doesn’t change anything.

This started happening when I switched from Private to Private with Storage and enabled background agents functionality. Not sure if related but nothing else was changed and then I observed this issue.

1 Like
3

Hi @MateuszTarnogorski @arturskruze thank you both for reporting this. More information is necessary to debug this issue.

Can you please try following and post more information:

  • Does it still occur on a fresh chat?
  • If using Background Agent, are there the requests slow as well? (you can click into the running process and see)
  • Which models are you using?
  • Can you post a Request ID where this occurred? (privacy disabled for that one request, but if possible also one if you have privacy with storage enabled)
  • Otherwise what date & time did it occur last so it can be compared to the usage log
4
  1. Yes
  2. On background agent it says "You need to increase your hard limit

Background Agent requires at least $10 remaining until your hard limit. Please increase your hard limit or enable usage-based pricing."

even tho i used 20$ and my limit is 100$.

  1. gpt-4, gpt4.1 Claude 3.7 thinking, claude 4 sonnet max

  2. 686fa7e6-77fc-4bd0-a843-cbf3a9c80c57

5

@deanrie is this “Slow request” from screenshot just a message about slow API responses or is it about the account using slow requests? Message is confusing

6

Slow request from screenshot is about account using slow requests. I do not use API

7

Cannot comment right now as the requests reset for me and now they are being used. This will probably start to happen again when I exhaust my available included fast requests.

  • Does it still occur on a fresh chat
    Yes, tried also fresh chat, closing and opening cursor app, etc.
  • If using Background Agent, are there the requests slow as well? (you can click into the running process and see)
    Cannot comment as I didn’t yet have it working as I have WSL and the code is stored on WSL directory, so it doesn’t see git there. Already sent a support request for this.
  • Which models are you using?
    It was happening for any model with MAX mode and sonnet 4. So I had to use gemini 2.5 without max to get anything working (with slow requests).
  • Can you post a Request ID where this occurred? (privacy disabled for that one request, but if possible also one if you have privacy with storage enabled)
    I know this was happening after I switched to privacy with storage.

Hope this helps with debugging.