Troubleshooting Gemini 2.5 Pro Model Errors

Support Bug Reports
1

image
image613×224 5.2 KB

Request ID: d2f3d0b8-30d5-4814-8b6b-50ed2305cb62
{“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}
ConnectError: [unavailable] Error
at yel.$endAiConnectTransportReportError (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:4918:251962)
at tcr._doInvokeHandler (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:493:20860)
at tcr._invokeHandler (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:493:20602)
at tcr._receiveRequest (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:493:19367)
at tcr._receiveOneMessage (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:493:18189)
at TSt.value (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:493:16283)
at ve._deliver (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:49:2962)
at ve.fire (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:49:3283)
at Mst.fire (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:4905:12154)
at d.onmessage (vscode-file://vscode-app/d:/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:7030:18167)

Hello, I’m still experiencing this bug when using the Gemini 2.5 Pro model… is there any solution?
I’ve already turned off all MCP servers. When the agent modifies a few lines of code, on the next turn this error appears.
T_T

2

hi @hhhnnnneeee could you check please, it seems like you have an API key or API URL configured?

3

Oh right, I forgot to revert the custom model configuration. I was using a custom Gemini API URL that I hosted myself with an OpenAI-compatible wrapper, so I override the OpenAI base URL for my custom model. I thought it would work, but I ran into a bug. Okay, I’ll just switch back to the default configuration instead.

4

Yes please let me know if using default config resolved it.

5

Okay, this works… so is this some kind of bug, or is it expected behavior when I use a custom model?

6

Its an issue with some API calls and some models, we are looking into that.