GPT-4 can produce graphical animations

Hi everyone!

I’ve built a simple prototype to generate animations with GPT-4 using Manim:

Manim is a Python library for creating complex animations. It’s the library used by channels like 3Blue1Brown and many more.

Since GPT-4 is getting better at programming. I was wondering what would happen if we connect the two. I call it Generative Manim.

I’ve tried simple things like “Draw a blue circle and turn it into a red square”, or “Draw a circle, and the write the text ‘This is a circle’”, and so on.

Succesful demos of the experiment (Video with examples): https://youtu.be/zjh_7cg6Aa4

This is a quick experiment to test GPT-4 in animations. An improved production-ready version would have many more things: a continuous error correction for certain parts, maybe a real-time editor, etc.

Try the prototype: https://generative-manim.streamlit.app

Repository: https://github.com/360macky/generative-manim

Flowchart of how it works:

3 Likes

Cool project, I’m a physics teacher with 0 understanding of coding but chatgpt is helping me a lot! I’m interested to see if this works for me, I’ve tried some animating with manim and was not disappointed.

Is it possible to feed you app with more information about what’s in the manim library?