I would like to report that the system is experiencing significant slowness when using the cursor/mouse, which is affecting the completion of daily tasks. We kindly request that this issue be checked and resolved as soon as possible to restore normal performance.
Please let me know if any additional information is needed.
Best regards,
Steps to Reproduce
Auto models
Operating System
Linux
Current Cursor Version (Menu → About Cursor → Copy)
Safe mode test: run cursor --disable-extensions from your terminal. Does the slowness persist?
A: yes
What kind of slowness you’re seeing (typing lag, UI freezes, etc.)? A: I am noticing delays in responses from the system — it seems like it is “thinking” for a few seconds before processing actions. The cursor and interface do not freeze.
Open Activity Monitor, sort by CPU/Memory, and capture a screenshot of Cursor processes A: I took a screenshot while the AI processing task was running in Cursor.
Additionally, I’ve noticed that the performance varies depending on the time of day. During late evening and night hours, the system becomes more usable and the slowness decreases. This gives the impression that the issue might be related to server overload or peak usage periods.
On MacOS, I’m noticing that Cursor seemingly has some control over my Edge browser, even though it’s supposed to only control a Google Chrome window it opens.
For example, if I navigate to a page in the in-app Browser Tab, my active Edge tab will navigate to the same page on my local dev site. If I play a YouTube video in either Edge or Cursor, it’ll play twice (because it starts playing in the other as well).
This has performance effects as well. While I have Cursor open, I basically cannot play YouTube videos in Edge. They get very choppy and pause to load a lot. Close Cursor, and the issue goes away.
EDIT: I’ve determined this is actually because I was running BrowserSync for my project without disabling Ghost Mode. Disabling it in my BrowserSync config fixed the performance/mirroring issue. Leaving this up in case it’s helpful to anyone else who makes the same mistake I did.
Just wanted to add to this — the experience has been quite frustrating lately. There are specific times of the day when the response speed from the model becomes extremely slow, to the point where it’s difficult to work efficiently.
It doesn’t seem related to local performance, since even in safe mode the delay persists. The lag feels more like a server-side or load-related issue, especially because performance noticeably improves late at night.
Would really appreciate if this could be investigated further — right now the slowdown is impacting productivity quite a bit.
