Introduction

This section explains how to use the VRM Desktop Pet.

First-Time Use

  1. You can simply click the "Start" button or "Start (Web)" button on the VRM Desktop Pet interface, and you will see the pet appear on your desktop or within the webpage.
  2. Left-click to rotate, scroll to zoom, and right-click to pan.
  3. The buttons on the pet interface, from top to bottom, are: Drag button (hold to drag the entire pet window); WS connection window (only when this button turns red can you attempt to reconnect the pet to the backend); Up and Down buttons (for switching VRM models); Refresh button (to reset the VRM model's position); Close button (to exit the pet).
  4. Enabling expressions will allow the LLM to output some emojis (which will not render in the chat window) to control the pet's expressions. Remember to turn this off when not using the pet.
  5. Asynchronous tools: Tool calls will not block the conversation, and the tool results will be added to the conversation after generation. Note! This only applies to streaming API calls and will not work with non-streaming API calls.
  6. Please upload a 3D model file in .vrm format, which must adhere to the VRM 1.0 standard, so you can have your own personalized desktop pet.
  7. In the chat interface, you can directly click the VRM Desktop Pet button above the input box to start the pet with one click!

Model Acquisition

  1. You can download VRoid Studio and create your own custom VRM model, similar to a character customization interface in games.
  2. You can also download VRM models made by others on VRoid Hub.
  3. The VRM ecosystem is very rich, with many useful software and platforms available for acquiring VRM models. Feel free to explore on your own.