I Built a YouTube Summarization Workflow in n8n — Automatically Fetches Videos & Emails AI Summaries

Built with n8n
1

Hi everyone, I wanted to share a workflow I built that dramatically changed how I learn from YouTube videos.

I love learning from YouTube — whether it’s podcasts, tutorials, or conference talks — but there’s never enough time to watch everything I want. I kept bookmarking videos and saying “I’ll watch this later,” but the backlog kept growing. That’s when I thought: why not automate the part that takes the most time — finding and summarizing the key takeaways?

So I used n8n to build a YouTube summarization automation that:

  1. grabs the latest videos from chosen channels,

  2. generates concise summaries and highlights using AI,

  3. sends a clean summary email to my inbox.

image
image1771×284 46.4 KB

Here’s how I did it — step‑by‑step.

:brick: Step 1 — Fetch Latest Videos via YouTube Data API

To start, I needed a reliable way to get new videos from YouTube channels without manually checking them. For that I used the YouTube Data API v3 – the official API for accessing video metadata like URLs, titles, and thumbnails.

:link: YouTube Data API overview: https://developers.google.com/youtube/v3/getting-started

:robot: Step 2 — Generate Summaries with an AI Video Summarization API

Next, I pass each new video link to WayinVideo’s video summarization API via an n8n HTTP Request node. The API automatically generates a concise summary for each video.

This step is the heart of the workflow. It’s fast, high-quality AI summaries, so that I can skip watching the full video and still capture the main points quickly.

:link: WayinVideo’s Video Summarization API Doc: https://wayin.ai/api-docs/video-summarization/

:brain: Step 3 — Format the Summary for Email

Once the API returns the summary, I used an n8n Function node to generate a clean HTML email. Each video section includes:

:check_mark: Title
:check_mark: Thumbnail
:check_mark: YouTube link
:check_mark: AI-generated summary and highlights with timestamps

:open_mailbox_with_raised_flag: Step 4 — Send Email with Summaries

Finally, the Email node (SMTP / Gmail) sends the compiled summaries to my inbox.

  • Set up the recipient address

  • Dynamic subject like: New YouTube Summaries — {{date}}

  • HTML body contains the formatted summaries

Now I get a readable digest instead of a long list of unwatched videos.

:hammer_and_wrench: Why This Workflow Works

:sparkles: Time-saving — No more watching 30–60 minute videos just to get the main ideas
:sparkles: Automated — runs on a schedule via n8n (e.g., every 6 hours)
:sparkles: High-quality summaries — AI handles the heavy lifting
:sparkles: Modular — easy to add new channels or change formatting

The key is combining n8n’s automation capabilities with a fast AI summarization step — without it, automating YouTube summaries would be much harder.

:link: Useful References

2

Great breakdown! One thing worth considering if you’re scaling this: WayinVideo’s API might have rate limits, so adding a delay node between video fetches is helpful if you’re processing many channels. Also, if you want better summaries for technical content, Claude (via n8n’s Anthropic integration) sometimes outperforms GPT for extracting key takeaways from dense talks. How many channels are you running this on?

3

This is a solid workflow! Using AI summarization + n8n automation for knowledge management is a perfect match. A few thoughts:

  1. For scaling this: If you’re planning to summarize high-volume channels, consider adding a deduplication step before the AI API (checking if the video already exists in your DB). Saves API costs big time.

  2. Timestamp extraction: You mentioned highlights with timestamps — if WayinVideo doesn’t provide those natively, you could feed the summary + transcript into Claude via the OpenAI/Anthropic API to extract key moments with timestamps. Makes the digest even more scannable.

  3. Weekly digest instead of 6-hour batches: For heavy channel subscriptions, a weekly summary-of-summaries (using another AI pass) reduces email fatigue while keeping content fresh.

  4. Workflow versioning: As you refine your summarization logic, version your n8n workflows in a Git repo — makes it easy to A/B test prompt changes and summaries side-by-side.

Nice implementation. This pattern (fetch → AI processing → formatted output) is incredibly useful for learning workflows. :+1: