workflows.fit
Back to n8n workflows
n8n templateFreeBy Adam Goodyer

Digest and analyse YouTube videos with Apify and Gemini Pro

Video Digestion Workflow — n8n Template Description --- How it works This workflow takes any YouTube video URL and automatically extracts a rich, structured analysis — including transcript, key visual moments, video m...

DevelopmentCore NodesAILangchainSticky NoteForm Trigger@apify/n8n-nodes-apify.apifyCode
Loading interactive preview...

Template notes

Video Digestion Workflow — n8n Template Description

---

How it works

This workflow takes any YouTube video URL and automatically extracts a rich, structured analysis — including transcript, key visual moments, video metadata, SEO keywords, and content section breakdowns. It's designed as the foundation layer for content repurposing, feeding its output into downstream workflows for creating Shorts, LinkedIn posts, Twitter threads, blog articles, email newsletters, and more.

The pipeline:

1. YouTube URL Input — A simple form trigger accepts any YouTube video URL. 2. Video Download (Apify) — Downloads the video file at 720p via the Apify YouTube Video Downloader actor. 3. Transcript Extraction (Apify) — Pulls the full transcript with timestamps from YouTube using the Apify YouTube Video Transcript actor. No audio processing needed — fast and reliable. 4. Data Consolidation — A Code node merges both Apify outputs into a single structured object containing: video URL, transcript text, timestamped segments, video metadata (title, description, duration, channel info, like/comment counts, thumbnail, publish date). 5. Visual Analysis (Google Gemini Pro) — Sends the actual video to Gemini's video analysis endpoint, which watches the entire video and identifies key B-roll moments with precise timestamps, app detection, and webcam overlay awareness. It categorises clips as clean screen recordings vs. webcam overlays vs. talking head segments. 6. Key Action Parsing — Filters and categorises the Gemini output into usable clips, removing talking-head-only segments and incomplete data. Outputs chronologically sorted clips with cropping metadata for downstream video editing. 7. AI Section Analysis (OpenAI) — Sends the transcript + key moments to OpenAI with structured output (JSON schema) to generate: video summary, one-liner, main argument, target audience, content style, tone, key takeaways, problems addressed, tools mentioned, frameworks explained, suggested titles, and SEO keywords. 8. Output — The final structured payload is ready to pass to any downstream workflow (e.g., Shorts creation, social media posting, blog generation).

Setup guide

Required accounts & API keys