# Code2Video: Video Generation via Code
Generate Educational video by Python manim code.
Yanzhe Chen,
Kevin Qinghong Lin,
Mike Zheng Shou
Show Lab @ National University of Singapore
📄 Paper |
🤗 Daily Paper |
🤗 Dataset |
🌐 Project Website |
💬 X (Twitter)
---
### Table of Contents
- [🌟 Overview](#-overview)
- [🚀 Quick Start: Code2Video](#-how-to-create----code2video)
- [1. Requirements](#1-requirements)
- [2. Configure LLM API Keys](#2-configure-llm-api-keys)
- [3. Run Agents](#3-run-agents)
- [4. Project Organization](#4-project-organization)
- [📊 Evaluation: MMMC](#-how-to-evaluate----mmmc)
- [🙏 Acknowledgements](#-acknowledgements)
- [📌 Citation](#-citation)
---
## 🌟 Overview
**Code2Video** is an **agentic, code-centric framework** that generates high-quality **educational videos** from knowledge points.
Unlike pixel-based text-to-video models, our approach leverages executable **Manim code** to ensure **clarity, coherence, and reproducibility**.
**Key Features**:
- 🎬 **Code-Centric Paradigm** — executable code as the unified medium for both temporal sequencing and spatial organization of educational videos.
- 🤖 **Modular Tri-Agent Design** — Planner (storyboard expansion), Coder (debuggable code synthesis), and Critic (layout refinement with anchors) work together for structured generation.
- 📚 **MMMC Benchmark** — the first benchmark for code-driven video generation, covering 117 curated learning topics inspired by 3Blue1Brown, spanning diverse areas.
- 🧪 **Multi-Dimensional Evaluation** — systematic assessment on efficiency, aesthetics, and end-to-end knowledge transfer.
---
## 🚀 Try Code2Video
### 1. Requirements
```bash
pip install -r requirements.txt
````
### 2. Configure LLM API Keys
Fill in your **API credentials** in `gpt_config.json`.
* **LLM API**:
* Required for Planner & Coder.
* Best Manim code quality achieved with **Claude-4-Opus**.
* **VLM API**:
* Required for Planner Critic.
* For layout and aesthetics optimization, provide **Gemini API key**.
* Best quality achieved with **gemini-2.5-pro-preview-05-06**.
* **Visual Assets API**:
* To enrich videos with icons, set `ICONFINDER_API_KEY` from [IconFinder](https://www.iconfinder.com/account/applications).
### 3. Run Agents
We provide two shell scripts for different generation modes:
#### (a) Any Query
Script: `run_agent_single.sh`
Generates a video from a single **knowledge point** specified in the script.
```bash
sh run_agent_single.sh --knowledge_point "Linear transformations and matrices"
```
**Important parameters inside `run_agent_single.sh`:**
* `API`: specify which LLM to use.
* `FOLDER_PREFIX`: output folder prefix (e.g., `TEST-single`).
* `KNOWLEDGE_POINT`: target concept, e.g. `"Linear transformations and matrices"`.
---
#### (b) Full Benchmark Mode
Script: `run_agent.sh`
Runs all (or a subset of) learning topics defined in `long_video_topics_list.json`.
```bash
sh run_agent.sh
```
**Important parameters inside `run_agent.sh`:**
* `API`: specify which LLM to use.
* `FOLDER_PREFIX`: name prefix for saving output folders (e.g., `TEST-LIST`).
* `MAX_CONCEPTS`: number of concepts to include (`-1` means all).
* `PARALLEL_GROUP_NUM`: number of groups to run in parallel.
### 4. Project Organization
A suggested directory structure:
```
Code2Video/
│── agent.py
│── run_agent.sh
│── run_agent_single.sh
│── api_config.json
│── ...
│
├── assets/
│ ├── icons/ # downloaded visual assets cache via IconFinder API
│ └── reference/ # reference images
│
├── json_files/ # JSON-based topic lists & metadata
├── prompts/ # prompt templates for LLM calls
├── CASES/ # generated cases, organized by FOLDER_PREFIX
│ └── TEST-LIST/ # example multi-topic generation results
│ └── TEST-single/ # example single-topic generation results
```
---
## 📊 Evaluation -- MMMC
We evaluate along **three complementary dimensions**:
1. **Knowledge Transfer (TeachQuiz)**
```bash
python3 eval_TQ.py
```
2. **Aesthetic & Structural Quality (AES)**
```bash
python3 eval_AES.py
```
3. **Efficiency Metrics (During Creating)**
* Token usage
* Execution time
👉 More data and evaluation scripts are available at:
[HuggingFace: MMMC Benchmark](https://huggingface.co/datasets/YanzheChen/MMMC)
---
## 🙏 Acknowledgements
* Video data is sourced from the **[3Blue1Brown official lessons](https://www.3blue1brown.com/#lessons)**.
These videos represent the **upper bound of clarity and aesthetics** in educational video design and inform our evaluation metrics.
* We thank all the **Show Lab @ NUS** members for support!
* This project builds upon open-source contributions from **Manim Community** and the broader AI research ecosystem.
* High-quality visual assets (icons) are provided by **[IconFinder](https://www.iconfinder.com/)** and **[Icons8](https://icons8.com/icons)**, which were used to enrich the educational videos.
---
## 📌 Citation
If you find our work useful, please cite:
```bibtex
@misc{chen2025code2videocodecentricparadigmeducational,
title={Code2Video: A Code-centric Paradigm for Educational Video Generation},
author={Yanzhe Chen and Kevin Qinghong Lin and Mike Zheng Shou},
year={2025},
eprint={2510.01174},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2510.01174},
}
```