Cursor infinite reconnecting .-

Support Help
1

Hello all, I’ve been running into a problem for two days now. The AI ​​model (Auto) is taking too long to reconnect and isn’t responding at all. Is there a solution?

image
image978×756 40.3 KB

3

Here are a few things you can try:

  • Switch models temporarily (pick a specific model instead of auto)
  • Restart Cursor (not just close the window—quit the app completely)
  • Log out and back in to refresh your auth session

Are you using a custom API key? What version of Cursor are you using?

5

Hey, I can see from your network diagnostics that DNS is extremely slow (12s instead of the normal <100ms), and all streaming connections (API, Ping, Chat, Agent) are getting dropped with ECONNRESET. Meanwhile, static endpoints (Marketplace, Auth, Tab) work fine.

This pattern usually points to something at the network level, like your ISP, a firewall, or a proxy doing Deep Packet Inspection and killing long-lived connections.

A few things to try:

  1. Mobile hotspot test: This is the fastest way to confirm it’s a network issue. Connect your PC to your phone’s mobile data (not your home WiFi) and try Cursor. If it works, the problem is your network or ISP.

  2. Try a VPN: If the hotspot works, a VPN may route around whatever is blocking the connections.

  3. Switch DNS: Your DNS lookup took 12 seconds, which is extremely slow. Try switching to Google DNS (8.8.8.8) or Cloudflare (1.1.1.1) in your OS network settings.

  4. Disable HTTP/2: Go to Cursor Settings > General, search for “HTTP/2”, then enable “Disable HTTP/2”. This can help on certain networks.

  5. Check firewall or antivirus: If you have any security software, try temporarily disabling it or adding exceptions for *.cursor.sh and *.cursor.com.

Can you also share:

  • What OS are you on?
  • What’s your ISP, and are you on a home or corporate network?
  • Are you using a VPN or proxy?

The hotspot test result will tell us the most. Let me know how it goes.

6

This has started happening to me too, same as OP: just started 2-3 days ago. The problem is flaky and not immediately reproducible.

Version: 2.5.26 (user setup)
VSCode Version: 1.105.1
Commit: 7d96c2a03bb088ad367615e9da1a3fe20fbbc6a0
Date: 2026-02-26T04:57:56.825Z
Build Type: Stable
Release Track: Default
Electron: 39.4.0
Chromium: 142.0.7444.265
Node.js: 22.22.0
V8: 14.2.231.22-electron.0
OS: Windows_NT x64 10.0.22631

NAME=“Amazon Linux”
VERSION=“2023”
ID=“amzn”
ID_LIKE=“fedora”
VERSION_ID=“2023”
PLATFORM_ID=“platform:al2023”
PRETTY_NAME=“Amazon Linux 2023.7.20250609”
ANSI_COLOR=“0;33”
CPE_NAME=“cpe:2.3:o:amazon:amazon_linux:2023”
HOME_URL=“xxxx”
DOCUMENTATION_URL=“xxxx”
SUPPORT_URL=“xxx”
BUG_REPORT_URL=“xxx”
VENDOR_NAME=“AWS”
VENDOR_URL=“https://aws.amazon.com/”
SUPPORT_END=“2029-06-30”

1 Like
7

Also having this issue this morning as well. Worked fine for an hour or so, and suddenly gets stuck on Reconnecting status. Using Opus 4.6 model primarily but I have tried other models. @deanrie

8

Hey @Kyle_Nguyen @glomar_dev,

I’m seeing the OP had a similar situation. DNS was slow, and streaming connections were dropping with ECONNRESET. You might have the same root cause, or it might be something else, so we should narrow it down.

Try these first:

  1. Disable HTTP/2: Cursor Settings > General, search for “HTTP/2”, enable “Disable HTTP/2”, then fully restart Cursor. In a similar thread this helped some users.

  2. Network Diagnostics: Cursor Settings > Network > Run Diagnostics, then paste the results here.

  3. Mobile hotspot test: Connect via mobile data. If it works there, the issue is likely your network.

@Kyle_Nguyen I see both Windows and Amazon Linux in your info. Are you using SSH Remote? If yes, that’s important context.

@glomar_dev The fact it worked for an hour and then failed could point to something at the network level killing long lived connections. Are you using a VPN, proxy, or a corporate network?

Let me know what you find, especially the network diagnostics output.

9

Hey Dean, sorry for the delay.

It’s working fine and well for me again today, both on and off company network, and on and off VPN (corporate). I don’t believe it was a network issue on my end, but who knows. All good now though, thanks for trying to help.

1 Like