Using Auto stop on model not working

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

When using “Auto mode”, it gives the error shown below.
But i thought auto mode would try another model if it doesn’t work?

Steps to Reproduce

When using “Auto mode”, it gives the error shown below.
But i thought auto mode would try another model if it doesn’t work?

Expected Behavior

If a model isn’t working in Auto Mode, it should try another model that we do have enabled.

Screenshots / Screen Recordings

Operating System

Windows 10/11

Current Cursor Version (Menu → About Cursor → Copy)

Version: 2.3.0-pre.6.patch.0 (user setup)
VSCode Version: 1.105.1
Commit: 37861b0694581c338e2dcb7caa5bc2127ded15c0
Date: 2025-12-13T17:18:26.627Z
Electron: 37.7.0
Chromium: 138.0.7204.251
Node.js: 22.20.0
V8: 13.8.258.32-electron.0
OS: Windows_NT x64 10.0.26200

For AI issues: which model did you use?

Auto

For AI issues: add Request ID with privacy disabled

Request ID: 1c506f95-6ba2-4556-949a-1b59ce59e1b4
{“error”:“ERROR_OPENAI”,“details”:{“title”:“Unable to reach the model provider”,“detail”:“We’re having trouble connecting to the model provider. This might be temporary - please try again in a moment.”,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:false}

Does this stop you from using Cursor

Yes - Cursor is unusable

Hey, thanks for the report. The ERROR_OPENAI message matches several recent cases.

Quick steps:

  • Settings > Network > Run Diagnostics - send full screenshot
  • Try disabling HTTP/2, restart Cursor
  • Try manually selecting a different model in the picker and see if it works
  • Check environment without extensions:
    • Run: cursor --user-data-dir="CursorData/TestProfile"
    • In the new window, test the same requests
  • If everything works in clean profile, check extensions
  • If using MCP servers, temporarily disable them and check again

Additionally, please share:

  • 2-3 fresh Request IDs from repeated errors
  • Logs from Help > Toggle Developer Tools > Console at the moment of error
  • Are you using VPN/proxy/Zscaler, have you tried mobile hotspot

Let me know what helped.

I did end up making a new chat, and referring back to the plan on the old chat, and then it did continue.
Not sure if it is related to the context used amount, if it would reach the limit or go over it, if that is why it stops.


  1. Will try with http/1
  2. Will try with another model.
  3. Will try cursor --user-data-dir=“CursorData/TestProfile”.
  4. Will try without any extensions enabled.
  5. I do not have any MCP’s enabled atm.
  6. I do not use a VPN/Proxy/Zscaler or Tailgate.
2 Likes

This topic was automatically closed 22 days after the last reply. New replies are no longer allowed.