Generate blog posts and tweet threads from YouTube with Supadata and OpenAI

Quick overview

This workflow collects a YouTube video URL via an n8n form, fetches the video transcript from Supadata (with async polling for longer videos), and uses OpenAI to generate a standalone blog post and a 5–8 tweet thread from the same transcript.

How it works

1. Receives a YouTube URL submitted through an n8n form trigger.
2. Requests the transcript from the Supadata Transcript API and continues even if the request returns an error.
3. If Supadata returns an async job ID, waits briefly and polls Supadata until the transcript job completes or fails.
4. Extracts the transcript text (or an error message) and truncates it to 8,000 characters.
5. Sends the transcript to OpenAI to write a 600–900 word blog post with a title, sections, and a closing takeaway.
6. Sends the same transcript to OpenAI to generate a numbered 5–8 tweet thread, keeping each tweet under 280 characters.

Setup

1. Create a Supadata account and add an HTTP Header Auth credential with header name x-api-key set to your Supadata API key, then select it in the Supadata transcript request and polling steps.
2. Add OpenAI credentials and ensure the OpenAI Chat Model node is authenticated.
3. Review and adjust the prompts for the blog post and tweet thread outputs to match your tone, length, and formatting preferences.