WebPNGTuber is an open-source application for creating and managing PNG Tuber models. It combines ease of use with advanced functionality, allowing you to build interactive avatars that react to your voice.
- 🎨 Multi‑layer editor: Build complex models from PNG and GIF images.
- 🔊 Audio reactions: Configure 4 reaction levels (silence, whisper, normal, shout).
- 👁️ Automatic blinking: Realistic eye animation with adjustable frequency.
- 🌐 Built‑in web server: Stream to OBS via
http://localhost:6969. - ⚙️ Advanced settings: Adjust sensitivity, noise reduction, effects.
- 💾 Slot system: Save up to 6 models with previews.
- 🌑 Idle mode: Automatically dim the avatar during inactivity.
The "Customizable Slugcat PNGTuber Model" is provided by CurioKryptic (itch.io):
https://curiokryptic.itch.io/slugcat-pngtuber-model
Terms: the model is free to use; use it for testing.
- Download the latest version from the Releases section.
- Extract the archive to any folder.
- Run
WebPNGTuber.exe.
# Clone the repository
git clone https://github.com/core-dump-dev/WebPNGTuber.git
# Install dependencies
pip install -r requirements.txt
# Run the application
python main.py- Windows 7 or newer
- For development: Python 3.7+
📦 Full list of dependencies (requirements.txt)
Pillow==10.3.0
numpy==1.26.4
sounddevice==0.4.6
Flask==3.0.3
requests==2.31.0- Start the web server inside the application.
- In OBS, add a new "Browser" source.
- Enter the URL:
http://localhost:6969. - Set the size to 700x700 pixels.
- Open the editor from the main window.
- Import PNG/GIF images.
- Arrange the layers in the desired order.
- Adjust position, scale, and rotation of elements.
- Group related elements (e.g., eyes).
- Configure reactions for different volume levels.
- Save the model to one of the 6 slots.
- Shake: Slight wobble on loud sounds.
- Bounce: Hopping animation.
- Pulse: Smooth size oscillation.
- Idle mode: Dimming when no sound is detected.
-
Closing the main window while saving a model
After saving in the editor, do not close it immediately – first check the model in the main window. -
Limited GIF support
For animations, use optimized files. -
Sound reaction delay
Make sure the correct audio device is selected in settings.
- Start simple: Body + mouth + eyes.
- Use transparency: PNGs with transparent backgrounds look better.
- Optimise GIFs: Reduce the number of frames and colours.
- Experiment: Try different reactions for different parts of the face.
- Save often: Regularly save your work into different slots.