n8n Workflow: Turn YouTube Transcripts into Newsletters & Social Posts

Turn any YouTube video into a ready-to-publish newsletter + social posts—automatically

This n8n workflow takes a YouTube URL (via webhook), grabs or accepts the transcript, uses OpenAI Chat Completions to generate a structured newsletter and repurposed content for X (Twitter) and LinkedIn, then returns the results as JSON (plus a rendered Markdown version).

What this workflow does

• Receives a POST webhook request with url (required) and optionally transcript.
• Validates the YouTube link by extracting the video ID from the provided URL (and errors if the URL is invalid).
• Obtains transcript text: uses the provided transcript if present; otherwise fetches captions from youtubetranscript.com and converts the transcript XML into plain text.
• Generates content with OpenAI: sends the transcript to the OpenAI Chat Completions API with instructions to return structured JSON including:
  • Subject line + preview text
  • Four newsletter sections
  • Takeaways
  • An X thread
  • A LinkedIn post
  • A call to action
• Parses and normalizes the response into consistent fields and adds video metadata (source URL and transcript word count).
• Builds a complete newsletter in Markdown (including the repurposed posts) and returns everything as a JSON response.

Use cases

• Repurpose educational YouTube content into a weekly newsletter
• Enable SaaS operators to publish faster by turning video transcripts into consistent, structured posts.
• Automation engineers use the webhook to integrate your existing content pipeline with n8n.

Technical details

• Nodes used: code, webhook, http request, respond to webhook
• Outbound requests: youtubetranscript.com for captions when no transcript is supplied
• AI generation: OpenAI Chat Completions API (replace YOUR_OPENAI_API_KEY in the Authorization header)
• Output: JSON payload + generated newsletter in Markdown

Note: You’ll need to deploy the webhook and configure your source system to POST { "url": "...", "transcript": "..." } (transcript optional).