AI Model Selector - VS Code Extension
A customizable AI model selector that supports OpenAI-compatible APIs, allowing you to use custom models in VS Code Chat.
Features
- OpenAI-compatible API support (DeepSeek, Qwen, MiMo, local models, etc.)
- Models automatically appear in VS Code Chat model picker
- Config file is watched for changes — just save and it takes effect
Usage
| Command |
Description |
ai-model: Config |
Open .vscode/model.json config file |
Configuration
Run the ai-model: Config command to open the config file, edit and save. The config file is located at .vscode/model.json.
{
"selectedModel": "deepseek-chat",
"models": [
{
"id": "deepseek-chat",
"name": "DeepSeek Chat",
"vendor": "deepseek",
"apiKey": "sk-your-api-key",
"baseUrl": "https://api.deepseek.com/v1",
"maxInputTokens": 128000,
"maxOutputTokens": 16000
}
]
}
Field Reference
| Field |
Description |
id |
Model ID used for API calls |
name |
Display name in VS Code Chat |
vendor |
Vendor identifier |
apiKey |
API key for authentication |
baseUrl |
API base URL (e.g. https://api.deepseek.com/v1) |
maxInputTokens |
Maximum input token count |
maxOutputTokens |
Maximum output token count |
Development
npm install
npm run compile
# Press F5 to launch debug
Package
npm install -g @vscode/vsce
vsce package
Commands
| Command |
Description |
modelSelector.config |
Open model configuration file |
License
MIT License
| |
