Unable verify openai key with selected model type

Bug Reports
1

clear description

Cannot verify the DeepSeek OpenAI key successfully because the “model” is always set to “default” instead of a custom model.

reproduce the bug

  1. Uncheck all models and add a new model named “deepseek-chat”.

    image
    image1159×1229 45.7 KB

  2. using valid key and base url according deepseek document

    image
    image1242×444 23 KB

  3. click verify and fail because the ‘model’ didnot change

    image
    image701×533 28.5 KB

command: noted that ** "model": "default" **

curl https://api.deepseek.com//chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer sk-fcd75ssssssssssssssss1e40f22c438d" -d "{ \"messages\": [ { \"role\": \"system\", \"content\": \"You are a test assistant.\" }, { \"role\": \"user\", \"content\": \"Testing. Just say hi and nothing else.\" } ], \"model\": \"default\" }"

error message:

{"error":{"message":"Model Not Exist","type":"invalid_request_error","param":null,"code":"invalid_request_error"}}

Cursor version

Version: 0.46.8 (user setup)
VSCode Version: 1.96.2
Commit: be4f0962469499f009005e66867c8402202ff0b0
Date: 2025-02-28T19:42:31.633Z
Electron: 32.2.6
Chromium: 128.0.6613.186
Node.js: 20.18.1
V8: 12.8.374.38-electron.0
OS: Windows_NT x64 10.0.26100

2 Likes
2

Hey, thanks for the report, it will be fixed soon.

3

I also have the same problem！

4

Same problem here. A new feature is needed to change the default “default” model name.

5

I have the same problem and I wonder when it will be fixed.
my version is 0.46.11

Version: 0.46.11 (user setup)
VSCode Version: 1.96.2
Commit: ae378be9dc2f5f1a6a1a220c6e25f9f03c8d4e10
Date: 2025-03-07T07:01:25.026Z
Electron: 32.2.6
Chromium: 128.0.6613.186
Node.js: 20.18.1
V8: 12.8.374.38-electron.0
OS: Windows_NT x64 10.0.19045

6

I had the same problem

7

me too。

版本: 0.46.11 (Universal)
提交: 1.96.2
日期: ae378be9dc2f5f1a6a1a220c6e25f9f03c8d4e10
Electron: 2025-03-07T06:44:15.844Z
ElectronBuildId: 32.2.6
Chromium: undefined
Node.js: 128.0.6613.186
V8: 20.18.1
OS: 12.8.374.38-electron.0

8

Is there any progress on this issue? Currently, in order to use custom models, I have to fallback to version 0.45.

Version 0.45 does not support the latest MCP features

9

Solved@0.47