🎬 AI Multi-Agent Video Creator

Multi-agent pipeline for creating longform HD videos with music, voiceover, and cinematic visuals

🌐 Live Demo → viral-video-maker.streamlit.app

---

A multi-agent video production pipeline. Enter a topic, configure the style, and let specialized AI agents handle every step: scriptwriting, visual scene generation, TTS narration, background music, and final video assembly — all the way to a downloadable MP4.

✨ Features

• Script Generation — Claude Sonnet writes a structured, narrated script
• Voiceover Narration — Minimax Speech O2 TTS with multiple voice options
• Visual Scene Creation — Luma Ray 2 generates cinematic video clips per scene
• Background Music — Google Lyria generates original background tracks
• Audio Mixing — optional voiceover + music blending
• Customizable Styles — Documentary, Cinematic, Nature, and more
• Aspect Ratios — 16:9, 9:16, 1:1, 4:3
• Video Lengths — 10s, 15s, or 20s
• Camera Motions — randomized zooms, pans, orbit, and more
• Downloadable Scripts — export the generated script separately

🚀 Quick Start

git clone https://github.com/RhythrosaLabs/multi-agent-viral-video-maker.git
cd multi-agent-viral-video-maker
pip install -r requirements.txt
streamlit run app.py

Or try the live demo.

🛠️ Tech Stack

• Python + Streamlit — web app and pipeline orchestration
• Anthropic Claude — script generation
• Replicate — Luma Ray 2 (video), Google Lyria (music)
• Minimax Speech — TTS narration
• MoviePy — video assembly and audio mixing

🤝 Contributing

PRs welcome. Open an issue first for major changes.

📄 License

MIT

💛 Support

If this pipeline helps you create content, consider supporting development:

👉 Donate via PayPal — @noodlebake

🌐 Portfolio: rhythrosalabs.github.io (more apps, music and sound design)

---

Built by Dan Sheils