# showlab/Code2Video

[ICML 2026] Video generation via code

Repository: https://github.com/showlab/Code2Video
Canonical: https://ross.abutalabs.com/products/code2video
Homepage: https://showlab.github.io/Code2Video/
Language: Python
License: MIT
License Family: permissive
Topics: coding, multi-agent, video-generation, education
Last push: 2026-08-24T11:00:47+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 35, longevity 24
- inputs: {"age_days": 338, "days_push": 9, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2003, forks 288 (observed 2026-08-28T04:06:04.498167+00:00)

## What it is
Code2Video is a research framework from Show Lab (NUS) that generates educational videos by having multi-agent LLM pipelines write and render code (e.g., Manim-style animations) rather than using diffusion video models. It targets precise, deterministic explainer videos for topics like math, algorithms, and neural networks.

## Use cases
- generate educational explainer videos from a topic
- create math or algorithm animation videos automatically
- use LLM agents to write animation code for teaching
- produce deterministic tutorial videos instead of diffusion-generated video
- make lecture-style videos for concepts like Fourier series or backpropagation

## When to choose
- you need precise, text-accurate educational videos that diffusion models render poorly
- you want programmatic, reproducible video generation via code
- you're researching code-centric video generation or multi-agent pipelines

## When to avoid
- you want photorealistic or cinematic video footage
- you need a polished end-user app rather than a research codebase
- you don't have access to capable LLM APIs to drive the agents

## Facets
- artifact type: library
- maturity: active
- function: video-processing, agent-framework, llm-inference
- domain: artificial-intelligence, education, large-language-models
- platform: python, cross-platform
- tags: educational-video-generation, code-as-video, multi-agent-llm, manim, research-project, icml-2026, code-generation, ai-agents, video

## Member repositories
- showlab/Code2Video (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:04.498167+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:01:55.782076+00:00, confidence not recorded.
  - readme: https://github.com/showlab/Code2Video (fetched 2026-08-28T04:06:04.498167+00:00, sha ac30ca0ed4d1)
  - homepage: https://showlab.github.io/Code2Video/ (fetched 2026-08-29T10:41:37.143185+00:00, sha 798ad97e3564)
- Data as of 2026-08-30T08:39:29.467469+00:00.