LM Studio Manager

Automatically loads models into LM Studio.

Installation

Install globally using npm or pnpm:

npm install -g lm-studio-manager
# or
pnpm add -g lm-studio-manager

Usage

1. Make sure LM Studio is running with WebSocket API enabled
2. Configure your models in models.config.json
3. Run the manager:

lm-studio-manager

Configuration

The models.config.json file should contain:

{
  "serverUrl": "ws://localhost:1234",
  "models": [
    {
      "identifier": "qwen/qwen3-coder-30b",
      "config": {
        "contextLength": 42000,
        "gpuOffload": "max",
        "keepModelInMemory": true,
        "flashAttention": true
      }
    }
  ]
}

Features

• Automatic model loading on startup
• Progress tracking during model loading
• Prevention of redundant model loading
• Support for both LLM and embedding models
• Configurable model parameters (context length, GPU offloading, etc.)