I added Azure OpenAI configuration in Cursor IDE.
Base URL, Deployment Name and API Key.
When I try to chat with Agent using GPT-5.5 I get:
Request ID: ac5d1512-3125-413a-82cb-84085eb645d5
{“error”:“ERROR_PROVIDER_ERROR”,“details”:{“title”:“Provider Error”,“detail”:“We’re having trouble finding the resource you requested. If the problem persists, please contact support.”,“isRetryable”:false,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:true}
Provider Error We’re having trouble finding the resource you requested. If the problem persists, please contact support.
kds: Provider Error We’re having trouble finding the resource you requested. If the problem persists, please contact support.
Hi @Limunce I apologize for the inconvenience. This is a known issue, and I’ve linked the forum post to the bug report so we can keep you updated on the status of it. In the meantime, the following options may immediately help you:
Use a non-GPT-5 model on your Azure key. Models outside the GPT-5 family — for example a GPT-4.1 or GPT-4o deployment — use Azure’s standard deployment endpoint and should work today with your existing BYOK configuration. Just deploy/select one of those in Azure and pick it in Cursor’s model picker.
If you specifically need GPT-5.5 in the meantime, you can use it through Cursor’s built-in model access (i.e., without the custom Azure key for that model). That path routes correctly. Note this draws on your normal Cursor usage rather than billing to your Azure account.
The Azure OpenAI BYOK issue with GPT-5 models, where you were hitting the “trouble finding the resource you requested” error, should have been fixed on our end in a recent deploy. Could you give it another try and let me know if you still run into it?
Hey thanks for lmk, ill test it when I have need, maybe you’ll be glad to know that with grok 4.5 release i havent even been using the gpt 5 models anymore!
hi its not working on the latest update i just downloaded 1 minute ago, wasn’t working before then either!:
Request ID: edec6fa3-f5a0-414a-8858-3d2e0e047d72
{“error”:“ERROR_PROVIDER_ERROR”,“details”:{“title”:“Provider Error”,“detail”:“We’re having trouble finding the resource you requested. If the problem persists, please contact support.”,“isRetryable”:false,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:true}
We’re having trouble finding the resource you requested. If the problem persists, please contact support.
Provider Error We’re having trouble finding the resource you requested. If the problem persists, please contact support.
NonRetriableError: Provider Error We’re having trouble finding the resource you requested. If the problem persists, please contact support.
at _Ob (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9323:6780)
at mBr (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9323:5417)
at ZHm (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9324:12870)
at Object.classify (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9324:17485)
at AOb (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9324:5856)
at async Kel.runWithSharedTurnRunner (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9324:16837)
at async YG.runAgentLoop (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:12036:52450)
at async LWl.streamFromAgentBackend (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:17886:15130)
at async LWl.getAgentStreamResponse (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:17886:24972)
at async P8e.submitChatMaybeAbortCurrent (vscode-file://vscode-app/c:/Users/gamin/AppData/Local/Programs/cursor/resources/app/out/vs/workbench/workbench.glass.main.js:9490:2511)
and naturally with the other URL:
Provider returned error: {“error”:{“code”:“BadRequest”,“message”:“API version not supported”}} Request ID: 7df190af-a141-4f37-a3f1-d3473d9c9f75
Hard to know exactly what URLs you’re talking about!
For the URL associated with this error:
Make sure the URL is nothing more than https://****.openai.azure.com (replace **** with your resource name) — no /openai/v1 or anything else on the end, Cursor appends the rest.
Also make sure the deployment field is set to the name of your actual Azure deployment, and that you’re pointing at the *.openai.azure.com resource rather than an Azure AI Foundry project endpoint (*.services.ai.azure.com), which we don’t currently support.
potentially flaky bug but super annoying, grok 4.5 says doesnt support custom api keys when ive enabled azure openai so is there a way i can keep azure enabled always + still use grok 4.5 directly?
Hey, small bug when using GPT-5.6 Limunce with Azure OpenAI: when I leave it to run autonomously, it works fine, but when I have to manually call /summarise, it says “context summarised”. It doesn’t actually spend any time summarising it, and I can see that the context hasn’t decreased. Bug: when I want to do manual /summaries, it doesn’t do it.
