N8N ViralFlow

A fully automated n8n-based pipeline that transforms user input into short, scroll‑stopping videos with hooks, dynamic visuals, and dramatic TTS narration and autmatically uploads it on their Youtube.

🛠️ Tech Stack

Example Youtube Channel

🎯 Table of Contents

• 📝 Summary
• ✨ Features
• 🛠️ Tech Stack
• 📋 Workflow Overview
• 🔧 n8n Nodes & AI Models
• 🚀 Quick Start
• 📫 Contact

📝 Summary

Users fill out a stylized “Brain Rotter” form with:

1. Topic (e.g. Indian history)
2. Duration (30s / 1min / 2min)
3. Generative Style (e.g. Disney Pixar)

The workflow then:

• Generates 20 facts via web search
• Writes a punchy script (“What if…?” hook, reveals, mic‑drop)
• Converts text→speech, splits into segments
• Creates scene prompts, generates images via Together.ai
• Transforms images→video clips, concatenates, captions, adds music
• Uploads final Short to YouTube with metadata & hashtags

✨ Features

• End‑to‑End Automation: From form input to published YouTube Short
• AI‑Powered Content: Facts generation, scriptwriting, image prompts
• Cinematic Visuals: Realistic/dark‑themed images in user‑chosen style
• Dynamic Audio: TTS narration + background music + word‑by‑word captions
• Short‑Form Focus: Maximum 59 s, mobile‑friendly, platform‑optimized

🛠️ Tech Stack

📋 Workflow Overview

1. Form Input Collection
   • formTrigger: Collects Topic, Duration, Style with custom CSS.
2. Facts Generation
   • langchain.agent (deepseek‑r1:8b): Fetches 20 trending, controversial facts.
3. Script Writing
   • langchain.agent (mistral:7b): Builds “SUPREME INDIANS” script with hook, reveals, pride, mic‑drop.
4. Script Cleanup
   • code: Strips formatting, newlines, asterisks.
5. Text‑to‑Speech
   • httpRequest → chatterbox‑tts‑server → returns WAV + timestamps.
6. Transcription & Segmentation
   • Transcribe audio → JSON/SRT.
   • Split into ≤2 s scenes (Split into N's Scenes, Fixer2).
7. Image Prompting
   • chainLlm (llama3.1): Generates cinematic prompts per segment.
8. Image Generation
   • httpRequest → Together.ai API → base64 PNG → convert & upload to S3.
9. Clip Creation & Assembly
   • httpRequest → transform image→video clip
   • combine → concatenate all clips
10. Enhancements
    • Captioning (word‑by‑word)
    • Background music layering (FFmpeg compose)
    • Trim to 59 s
11. YouTube Upload
    • youTube node: Publishes as a public Short with title, description, tags.

🔧 n8n Nodes & AI Models

• n8n-nodes-base.formTrigger – Custom CSS form
• @n8n/n8n-nodes-langchain.agent – deepseek‑r1:8b & mistral:7b
• @n8n/n8n-nodes-langchain.chainLlm – llama3.1 for image prompts
• n8n-nodes-base.httpRequest – TTS, transcription, image & video services
• n8n-nodes-base.code – JS for cleaning, splitting, merging, random zoom speeds
• n8n-nodes-base.s3 – MinIO/AWS S3 uploads
• n8n-nodes-base.aggregate / merge / splitInBatches – Data orchestration
• n8n-nodes-base.youTube – YouTube Shorts publishing

🚀 Quick Start

Learn how to install and get started with each component:

• n8n Self Hosted AI – A Must Try GitHub Open Source Project!
  https://youtu.be/OCet5rGwEeI?si=sOSSSR8-QbAhUqsM

• Chatterbox TTS Server – Setup & Demo
  https://youtu.be/VEEn6h2aRR0?si=c-8Lo1YG-iJAw5G8

• SearchXng – Installation & Configuration
  https://youtu.be/O5b9FvHYjqc?si=MOKRk68Gm12IzS8M

• Workflow Inspiration – Automated Content Pipeline Overview
  https://youtu.be/sOylPpFyQ8A?si=EfouFqOt1n0UZgqO

• Ngrok – Host N8N for free
  https://youtu.be/RvAD2__YYjg?si=3_6tqVncQ0e5yqwL

📫 Contact

• Made by Prabhanshu Kamal
• GitHub: PrabhanshuKamal2121
• LinkedIn: PrabhanshuK

⚖️ Disclaimer

• This pipeline is for educational/demo purposes. It uses publicly available AI models and services. Results may vary; always validate before publishing.