Chibikko

Meet Chibikko, a tiny cuddly creature lives on your desktop!

Build

Prerequisite

Live2d Model

Before seeing the little creature, you need to download the live2d model

Free models are available on (live2d)[https://www.live2d.com/en/learn/sample/]

1. Click the Download Pro version for Shizuku to download the model
2. Unzip the model and put the unzipped folder into chibikko/public/

Live2d Cubism SDK

The live2d model relies on the (Live2D Cubism SDK)[https://www.live2d.com/sdk/download/web/]

Download the SDKs and put them under chibikko/public/scripts/

The file tree of the chibikko/public/ should look like:

chibikko\public
├─scripts
|    └─live2dcubismcore.js
|    └─live2dcubismcore.js.map
|    └─live2dcubismcore.min.js
└─shizuku

Voice

To make Chibikko speak in a voice you like!

The speak part is supported by (GPT-Sovits)[https://github.com/RVC-Boss/GPT-SoVITS].

To deploy GPT-Sovits, follow the instructions below:

1. Install (miniconda)[https://www.anaconda.com/docs/getting-started/miniconda/install]
2. Install the dependencies

    conda create -n GPTSoVits python=3.10
    conda activate GPTSoVits
    # Select the option, an example already tested on RTX5060 is
    # .\install.ps1 -Device CU126 -Source ModelScope
    pwsh -F install.ps1 --Device <CU126|CU128|CPU> --Source <HF|HF-Mirror|ModelScope> [--DownloadUVR5]

3. Running the GPT-SoVits

    python api.py

4. Prepare the voice you like by putting an example.wav file under the path C:\Users\{USER}\AppData\Roaming\\com.mashirocl.chibikko\ of that voice and also the text of that example.wav.
5. Prepare config file by creating a file named app_config.json like below:

{
  "ollm_model": "gemma3:4b",
  "tts_api_endpoint": "http://127.0.0.1:9880",
  "tts_example_wav":"example.wav",
  "tts_example_wav_text": "text"
}

Start

$ cd chibikko
$ npm install
$ npm run tauri dev