Gemini 3 API error

[mulgogi]

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

I’m using gemini3 via API. It takes a very long time to load and then an error occurs.

Network diagnosis is normal. There is no problem.

It was working fine about an hour ago. Looking at the API logs, I can see that requests are coming in.

Steps to Reproduce

You can select and use genimi3 in Agent Chat.

Operating System

MacOS

Current Cursor Version (Menu → About Cursor → Copy)

Version: 2.1.32
VSCode Version: 1.105.1
Commit: ef979b1b43d85eee2a274c25fd62d5502006e420
Date: 2025-11-25T05:00:12.749Z
Electron: 37.7.0
Chromium: 138.0.7204.251
Node.js: 22.20.0
V8: 13.8.258.32-electron.0
OS: Darwin arm64 25.1.0

For AI issues: which model did you use?

gemini3pro

For AI issues: add Request ID with privacy disabled

First error

We encountered an issue when using your API key: Provider was unable to process your request
API Error:

Request failed with status code 503: [{
  "error": {
    "code": 503,
    "message": "The model is overloaded. Please try again later.",
    "status": "UNAVAILABLE"
  }
}
]

Request ID: e80877a2-99b6-4fa1-a3e8-7fca72b5b042
{"error":"ERROR_OPENAI","details":{"title":"Unable to reach the model provider","detail":"We encountered an issue when using your API key: Provider was unable to process your request\n\nAPI Error:\n\n```\nRequest failed with status code 503: [{\n  \"error\": {\n    \"code\": 503,\n    \"message\": \"The model is overloaded. Please try again later.\",\n    \"status\": \"UNAVAILABLE\"\n  }\n}\n]\n```","additionalInfo":{},"buttons":[],"planChoices":[]},"isExpected":true}
ConnectError: [invalid_argument] Error
    at EZc.$endAiConnectTransportReportError (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6331:408452)
    at GPo._doInvokeHandler (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:21873)
    at GPo._invokeHandler (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:21615)
    at GPo._receiveRequest (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:20377)
    at GPo._receiveOneMessage (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:19194)
    at yPt.value (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:17286)
    at _e._deliver (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:49:2962)
    at _e.fire (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:49:3283)
    at Xmt.fire (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6316:12156)
    at MessagePort.<anonymous> (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:8973:18439)

Second error

We encountered an issue when using your API key: Provider was unable to process your request

API Error:The HTTP/2 stream has been early terminated

Request ID: 7bb05e83-aa49-4b32-ab70-366a5bbd0cd0
{"error":"ERROR_OPENAI","details":{"title":"Unable to reach the model provider","detail":"We encountered an issue when using your API key: Provider was unable to process your request\n\nAPI Error:\n\n```\nThe HTTP/2 stream has been early terminated\n```","additionalInfo":{},"buttons":[],"planChoices":[]},"isExpected":true}
ConnectError: [invalid_argument] Error
    at EZc.$endAiConnectTransportReportError (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6331:408452)
    at GPo._doInvokeHandler (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:21873)
    at GPo._invokeHandler (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:21615)
    at GPo._receiveRequest (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:20377)
    at GPo._receiveOneMessage (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:19194)
    at yPt.value (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6948:17286)
    at _e._deliver (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:49:2962)
    at _e.fire (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:49:3283)
    at Xmt.fire (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:6316:12156)
    at MessagePort.<anonymous> (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:8973:18439)

Does this stop you from using Cursor

Yes - Cursor is unusable

[deanrie]

Hey, thanks for the report. You’re running into two different issues:

1. The 503 “model overloaded” error comes directly from the Google Gemini API, not from Cursor. This means Google’s servers are overloaded. It’s a temporary issue on Google’s side, so you’ll need to wait and try again later, or temporarily switch to another model.

2. The HTTP/2 stream termination error is most likely related to recent changes in the Gemini API. Please try this:

   • Open Cursor Settings CMD+,
   • Search for “HTTP/2”
   • Enable “Disable HTTP/2”
   • Restart Cursor

Workaround options:

• Try a different Google AI model if your API key has access to multiple models
• Temporarily switch to another model provider
• Try again a bit later, when Google’s API load may be lower

If the issues are still happening after a few hours, please send an updated Request ID.