Home / Directory / AI video agents
Open-source agents and automation projects that turn an idea, topic or script into AI video. Each one is checked on GitHub with its license and last commit date.
8 verified entries. We list a project only after checking its repository or the vendor's own page. Install snippets are copied from the official README or docs, and summaries are in our own words. Last checked 2026-10-16. Know one we missed? Tell us.
Filter these in the full directory →
By HKU Data Intelligence Lab · License: MIT · Last commit 2026-09-30 · ★ 12,674
A research agent system that splits video making into roles (director, screenwriter, producer and generator) and goes from an idea or script to a multi-shot video while keeping characters consistent.
git clone https://github.com/HKUDS/ViMax.git
cd ViMax
uv syncRepository ↗ · Checked 2026-10-16
By harry0703 · License: MIT · Last commit 2026-10-16 · ★ 132,121
A popular open-source pipeline that makes a short vertical video from a topic or keyword. It writes the script, gathers stock footage, adds voiceover and subtitles, and renders the clip.
python -m pip install mpt
mpt webuiRepository ↗ · Checked 2026-10-16
By HBAI · License: MIT · Last commit 2026-10-13 · ★ 17,973
A self-hosted studio for short dramas and motion comics, with script writing, storyboards, character and scene assets, and image and video generation on an infinite canvas with AI agents. It supports MCP and plugins. The docs are mostly in Chinese.
git clone https://github.com/HBAI-Ltd/Toonflow-app.git
cd Toonflow-app
docker compose up -d --buildRepository ↗ · Checked 2026-10-16
By aqm857886159 · License: AGPL-3.0 · Last commit 2026-10-16 · ★ 577
A local-first video workbench that coding agents (Claude Code, Codex, Cursor) can direct over MCP. It covers storyboard, references, generation with any model or local ComfyUI, and an editable first cut on a timeline.
Works with: ComfyUI
git clone https://github.com/aqm857886159/Nomi.git
cd Nomi
corepack enable
pnpm install
pnpm devRepository ↗ · Checked 2026-10-16
By neopen · License: MIT · Last commit 2026-10-14 · ★ 225
An agent that breaks a script, novel or screenplay into shots and writes a consistent text-to-video prompt for each, ready for Sora, Veo or Runway. You can call it through MCP or a REST API. The docs are mostly in Chinese.
Works with: Sora, Veo, Runway
pip install penshotRepository ↗ · Checked 2026-10-16
By Agents365-ai · License: MIT · Last commit 2026-09-14 · ★ 1,682
A skill for coding agents that turns a topic into a narrated video built with Remotion, with local text-to-speech and cost limits on any AI generation.
npx skills add Agents365-ai/video-podcast-maker/skills/video-podcast-maker -gRepository ↗ · Checked 2026-10-16
By RayVentura · License: MIT · Last commit 2025-02-10 · ★ 8,035
An experimental framework for automating YouTube Shorts and TikTok channels, covering script, voiceover, footage and editing. It hasn't been updated since February 2025.
docker build -t short_gpt_docker:latest .
docker run -p 31415:31415 --env-file .env short_gpt_docker:latestRepository ↗ · Checked 2026-10-16
By FujiwaraChoki · License: MIT · Last commit 2026-03-26 · ★ 14,061
The original small project that automates YouTube Shorts with MoviePy, going from a topic to a finished short clip.
uv pip install -U wheel
uv pip install -U playsoundSnippet is the README's dependency step; follow the README for the full setup.
Repository ↗ · Checked 2026-10-16
Get the free guide
Getting Paid for AI Video: 7 paths, published pay figures with sources, and how to land a first client. No income promises.
We store your email to know who downloaded the guide. No spam. Privacy.