Config VS Code (Insiders): Difference between revisions

From Master of Neuroscience Wiki
Created page with "For the foreseeable future you need to install the VS Code '''''<u>Insiders</u>''''' version!!! See [https://code.visualstudio.com/insiders/ here] Ctrl+Shift+P : Preferences: Open User Settings (JSON) This is an example. Please note the "qwen3-32" selects the model not "name": "qwen3-32"<syntaxhighlight lang="json"> { "github.copilot.chat.customOAIModels": { "qwen3-32b": { "name": "qwen3-32b", "url": "https://chat-ai.academiccloud.d..."
 
No edit summary
 
(2 intermediate revisions by one other user not shown)
Line 1: Line 1:
For the foreseeable future you need to install the VS Code '''''<u>Insiders</u>''''' version!!! See [https://code.visualstudio.com/insiders/ here]
For the foreseeable future you need to install the VS Code '''''<u>Insiders</u>''''' version!!! See [https://code.visualstudio.com/insiders/ here]


Ctrl+Shift+P : Preferences: Open User Settings (JSON)
Ctrl+Shift+P: Preferences: Open User Settings (JSON)
 
 


This is an example. Please note the "qwen3-32" selects the model not "name":  "qwen3-32"<syntaxhighlight lang="json">
This is an example. Please note the "qwen3-32" selects the model not "name":  "qwen3-32"<syntaxhighlight lang="json">
Line 9: Line 11:
             "name": "qwen3-32b",
             "name": "qwen3-32b",
             "url": "https://chat-ai.academiccloud.de/v1",
             "url": "https://chat-ai.academiccloud.de/v1",
             "toolCalling": false,
             "toolCalling": true,
             "vision": false,
             "vision": false,
             "thinking": true,
             "thinking": true,
Line 19: Line 21:
     "workbench.startupEditor": "none"
     "workbench.startupEditor": "none"
}
}
</syntaxhighlight>
</syntaxhighlight>Next
 
Ctrl + Shift + P: Chat: Manage Language Models
[[File:2025-11-03 18-18.png|800x800px]] 
 
If "OpenAI Compatible" is not listed, you need to click "+ Add Models..." and select the option. The setting icon of "OpenAI Compatible" allows you to set the API key. It will not be present, if a model of this type is running.
 
I had to press the non hide button in the front of the model name.
 
I closed and opened VS Code. Now I can use that model.
 
Example interaction:
 
[[File:2025-11-03 18-40.png]]

Latest revision as of 09:54, 6 November 2025

For the foreseeable future you need to install the VS Code Insiders version!!! See here

Ctrl+Shift+P: Preferences: Open User Settings (JSON)


This is an example. Please note the "qwen3-32" selects the model not "name": "qwen3-32"

{
    "github.copilot.chat.customOAIModels": {
        "qwen3-32b": {
            "name": "qwen3-32b",
            "url": "https://chat-ai.academiccloud.de/v1",
            "toolCalling": true,
            "vision": false,
            "thinking": true,
            "maxInputTokens": 128000,
            "maxOutputTokens": 8192,
            "requiresAPIKey": true
        }
    },
    "workbench.startupEditor": "none"
}

Next

Ctrl + Shift + P: Chat: Manage Language Models

If "OpenAI Compatible" is not listed, you need to click "+ Add Models..." and select the option. The setting icon of "OpenAI Compatible" allows you to set the API key. It will not be present, if a model of this type is running.

I had to press the non hide button in the front of the model name.

I closed and opened VS Code. Now I can use that model.

Example interaction: