🎬 Generate cinematic videos from text prompts with GPT-4o, Fal.AI Seedance & Audio

⚡ 24,550 views · 🎬 Content Creation & Video

💡 Pro Tip — Since TikTok doesn’t have a public API, getting data usually means fragile HTTP scraping. ScraperNode is a community node that gives you clean JSON for TikTok profiles, videos, and comments — drop it into your workflow and skip the parsing.

Description

Who’s it for?

This workflow is built for:

• AI storytellers, content creators, YouTubers, and short-form video marketers
• Anyone looking to transform text prompts into cinematic AI-generated videos fully automatically
• Educators, trainers, or agencies creating story-based visual content at scale

What It Does

This n8n workflow allows you to automatically turn a simple text prompt into a multi-scene cinematic video, using the powerful Fal.AI Seedance V1.0 model (developed by ByteDance — the creators of TikTok).

It combines the creativity of GPT-4o, the motion synthesis of Seedance, and the automation power of n8n to generate AI videos with ambient sound and publish-ready format.

How It Works

1. Accepts a prompt from Google Sheets (configurable fields like duration, aspect ratio, resolution, scene count)
2. Uses OpenAI GPT-4o to write a vivid cinematic narrative
3. Splits the story into n separate scenes
4. For each scene:
   • GPT generates a structured cinematic description (characters, camera, movement, sound)
   • The Seedance V1.0 model (via Fal.AI API) renders a 5s animated video
   • Optional: Adds ambient audio via Fal’s MM-Audio model
5. Finally:
   • Merges all scene videos using Fal’s FFmpeg API
   • Optionally uploads to YouTube automatically

Why This Is Special

• Fal.AI Seedance V1.0 is a highly advanced motion video model developed by ByteDance, capable of generating expressive, stylized 5–6 second cinematic clips from text.
• This workflow supports full looping, scene count validation, and wait-polling for long render jobs.
• The entire story, breakdown, and scene design are AI-generated — no manual effort needed.
• Output is export-ready: MP4 with sound, ideal for YouTube Shorts, Reels, or TikTok.

Requirements

• n8n (Self-hosted recommended)
• API Keys:
  • Fal.AI (https://fal.ai)
  • OpenAI (GPT-4o or 3.5)
  • Google Sheets Example Google Sheet

How to Set It Up

1. Clone the template into your n8n instance
2. Configure credentials:
   • Fal.AI Header Token
   • OpenAI API Key
   • Google Sheets OAuth2
   • (Optional) YouTube API OAuth
3. Prepare a Google Sheet with these columns:
   • story (short prompt)
   • number_of_scene
   • duration (per clip)
   • aspect_ratio, resolution, model
4. Run manually or trigger on Sheet update.

How to Customize

• Modify the storytelling tone in GPT prompts (e.g., switch to fantasy, horror, sci-fi)
• Change Seedance model params like style or seed
• Add subtitles or branding overlays to final video
• Integrate LINE, Notion, or Telegram for auto-sharing

Example Output

Prompt: “A rabbit flies to the moon on a dragonfly and eats watermelon together”
→ Result: 3 scenes, each 5s, cinematic camera pans, soft ambient audio, auto-uploaded to YouTube Result

🔗 Nodes Used

Google Sheets, HTTP Request, YouTube, AI Agent, OpenAI Chat Model, Structured Output Parser

📥 Import

Download workflow.json and import into n8n: Workflow menu → Import from File

📖 Importing guide · 🔑 Credential setup