Our take
- Activity
- 19
- Community
- 19
- Issues
- 21
- Pull requests
- 20
+15 stars in the last 30 days
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
3,534Forks
225Change since Nov 17, 2026: 0 stars·0 forks
Overview
ManimML is a Python library that adds reusable machine learning visualizations and animations to the Manim Community Library. It helps educators, students, and video creators turn concepts such as neural network layers and forward passes into scenes, without building every visual element and animation from scratch.
Use it when you want editable, code-driven animations for teaching or explaining machine learning. It is a visualization tool, not a machine learning training framework.
Key Features
- Compose neural networks from feed-forward, convolutional, image, and max-pooling layers.
- Arrange layers into a network with configurable spacing and layer parameters.
- Generate animations of neural network forward passes.
- Visualize activation functions such as ReLU and Sigmoid on supported layers.
- Create dropout animations, optionally including a forward pass.
- Use Manim scenes and Python code to customize and combine visualizations.
Use Cases
- Instructors can create animated explanations of neural network architectures for lectures.
- Video creators can build code-driven visual sequences about convolution, pooling, or forward passes.
- Students can make visual supplements to coursework when a static diagram is not enough.
- Researchers and technical communicators can illustrate model components for presentations or educational materials.
Getting Started
Install Manim Community Edition first, then install ManimML:
pip install manim_ml
See the README for scene examples and rendering instructions. The README notes that some recent features may require installing from source.
Alternatives
- manim: Manim is a general-purpose animation engine for building explanatory visuals, while ManimML provides reusable components specifically for neural networks and machine-learning operations.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| ManimML | Python | MIT | 3.5k | Not checked yet |
| manim | Python | MIT | 94.5k | Active |
Considerations
ManimML requires the Manim Community Library and familiarity with Python-based Manim scenes. The README describes the project as under development and notes that documentation is incomplete and some features may change. Some recent features may not be available in the package installation and may require a source install. It provides visualization components, not model training or inference functionality.
Found this useful?
Share it with someone who would like ManimML.