Cursor has been stuck on “Planning next moves” for an extended period and I haven’t been able to resolve it. I’ve seen similar reports from other users and tried the recommended troubleshooting steps, but the issue persists.
What I’ve already tried:
• Uninstalled/reinstalled Cursor (including downgrading: removed the newer version and installed v2.0)
• Restarted Cursor multiple times and rebooted my computer multiple times
• Network connectivity checks
• Cleared/disabled extensions and extension data
• Switched models (Claude Sonnet → GPT-5.2 Agent)
• Other general debugging steps, with no change
This has been happening since about two weeks ago, and it’s blocking my work. Could you please help me diagnose and fix this?
Steps to Reproduce
Cursor version:2.3.34 (user setup) on Windows x64. when I type a question in the chat window, it stays stuck on “Planning next moves” for hours
Operating System
Windows 11
Current Cursor Version (Menu → About Cursor → Copy)
Hey, thanks for the report. This is a known issue and there are a few possible causes. I can see you’ve already tried a lot of things, but for debugging we’ll need a bit more info. Let’s go step by step:
Network Diagnostics (first)
Settings > Network > Run Diagnostics. If anything fails (DNS, HTTP, SSL), that could be the reason. Are you using a VPN, a corporate network, or antivirus? They often block HTTP/2.
Try disabling HTTP/2
Settings > Network > HTTP Compatibility Mode > HTTP/1.1. This is especially relevant for corporate networks, VPNs, or Zscaler.
Fully disable extensions
Run from the command line: cursor --disable-extensions. If Chat works, then an extension is causing it. Enable them one by one to find the culprit.
Test without a project
Close Cursor, then open it WITHOUT any project (empty window) and try Chat. If it works, the issue is likely related to your project indexing.
Same here. I wish I could provide a solution, but something that I noticed was that the agent would not respond until my codebase was indexed. I only have 48 files, but every time I disconnect from my Raspi and reconnect, the indexing starts again and just goes on forever.
Hi, I experience the same issue since yesterday. I have an ongoing plan that was interrupted in the middle, and I can’t resume it. It’s stuck at “planning moves”.
I have no network issue when running network diagnostics. Switching from HTTP/2 to HTTP/1.1 didn’t change anything.
I have disabled all extensions, but there’s no change (below screenshot is done with extensions disabled).
I’m on a small project, and it indexed 187 files successfully. I don’t see any indexing issue in indexing settings.
Then I opened a new window without project, and I raised a dummy question with a web search. It processed it successfully.
Here is my request ID: d16e897e-9560-4d9a-b05f-82cf081c267a
This morning, I upgraded to 2.4.21 (I’m an enterprise user and don’t have access to beta releases). Looks like it unblocked my issue, I could resume the plan implementation.
