Cursor’s chat freezes every so often and it forces me to close and re open the application. I am not sure why this happens. I have reinstalled everything multiple times already.
Hi @Senhua ,
I am just a fellow user, but perhaps the following information from you would be helpful:
System Information
Cursor
> Help
> About
Interface Details
Confirm which of the Cursor AI interfaces you are having trouble with:
Ctrl
+L
- Normal ChatCtrl
+L
- Long Context ChatCtrl
+L
- Interpreter ModeCtrl
+K
- In EditorCtrl
+K
- In TerminalCtrl
+I
(Composer)
LLM Details
Specify which LLM you are using:
cursor-small
gpt-3.5
gpt-4o-mini
gpt-4o
claude-3.5-sonnet
gpt-4o-128k
gemini-1.5-flash-500k
claude-3-haiku-200k
claude-3.5-sonnet-200k
Symbols Usage
Are you using any of the @
symbol features?
Chat Context
Which button are you choosing in the chat interface:
- Chat with context
- Chat with no context
- Chat with codebase
Relevant Logs
Check for any relevant error messages or logs:
Cursor
> Help
> Toggle Developer Tools
(and then click the Console
tab)
Also see:
01) Ctrl
+ Shift
+ P
(open Command Palette)
02) Type Developer: Open Log
and select Developer: Open Log File...
03) Type Cursor
and you will see some relevant options
If you experience any crashes, you can refer to this post for more information:
Normal chat
happens with gpto and 3.5 sonnet, but I also use mini sometimes
yes I frequently use the @ symbol feature to insert code for context
Version: 0.39.4
VSCode Version: 1.91.1
Commit: 0c8eb0f46a532a73f16f4427fcdaf7cd9070e140
Date: 2024-08-10T05:38:34.716Z
Electron: 29.4.0
ElectronBuildId: undefined
Chromium: 122.0.6261.156
Node.js: 20.9.0
V8: 12.2.281.27-electron.0
OS: Windows_NT x64 10.0.22631
I do get this error sometimes when the chat freezes, something about an operation being aborted
I found a workaround for this that I often use in ChatGPT: simply start a new chat to stop the freezes.
This issue often occurs when you have a long chat history. It seems that the AI needs to quickly process the entire conversation each time before responding to a new prompt. When the history gets too long, it takes longer for the model to process, and the more advanced the model, the longer it takes—sometimes causing it to time out.
If this works similarly to ChatGPT, there may come a point when opening a new chat no longer fixes the problem. In that case, you would need to delete older chats to free up memory.
Hope this helps.