Anim-Vid-AI

Turn boring math/algorithms into engaging animation videos

Comment

Theme: Educational Tools & Platforms / Artificial Intelligence

🚀 Vision & Purpose

Complex logic is hard to explain with just text. Whether it's the swapping logic of Bubble Sort, the traversal path of DFS (Depth First Search), or abstract Mathematical proofs, static images often fail to convey the movement of logic.

Anim-Vid-AI is an AI-powered visualization engine that turns natural language descriptions into high-quality, programmatic animations. It democratizes the power of the Manim engine (used by 3Blue1Brown), allowing developers, educators, and content creators to generate complex algorithm visualizations in seconds without writing a single line of Python code.

✨ Key Features

  • Algorithm Visualization: specialized in rendering sorting algorithms, graph traversals, and data structures.
  • Logic-to-Code: Converts prompts like "Show a binary search tree balancing itself" into precise Manim Python code.
  • Smart Iteration: Use the chat interface to refine animations (e.g., "Slow down the sorting step" or "Highlight the pivot element").
  • Dual View: targeted at developers—view the generated Python code alongside the rendered video to learn Manim syntax.
  • Production Ready: Dockerized architecture solving the complex dependency issues of FFmpeg, LaTeX, and Cairo.

🎥 Demo

Live URL: https://anim-vid-ai.onrender.com

Demo Video: https://www.youtube.com/watch?v=55oR1rhxszI

Github link: https://github.com/ovi054/anim-vid-ai

🧠 How It Works

  1. Intent Parsing: The AI analyzes the prompt to identify if the request is Mathematical (Graphs, Calculus) or Algorithmic (Arrays, Nodes, Pointers).
  2. Code Synthesis: It generates a MyScene class in Python, utilizing Manim's VGroup, Arrow, and .animate methods to represent logic flows.
  3. Compilation: The code is executed in an ephemeral sandbox to render the .mp4 output.
  4. Error Correction: If the code fails (e.g., overlapping nodes), the "Edit" feature allows the AI to self-correct the spatial arrangement.

🛠️ Tech Stack

  • Core Engine: Python, Manim Community (v0.19.0)
  • Intelligence: Fal AI (Claude 4.5 Sonnet) with custom prompting for algorithmic logic.
  • Interface: Gradio (Web UI).
  • Infrastructure: Docker & Render (Cloud Deployment).

Built With

Share this project:

Updates