A solo SaaS idea for turning long YouTube videos into short clips

A working tool turns long YouTube videos into short vertical clips from start to finish. The flow downloads the video with yt-dlp, turns speech into text with Whisper, lets Gemini choose the best moments, lets the user approve clips in a web editor, and then uses ffmpeg to render clips with captions for download. The current version runs on a personal computer as an n8n workflow, using command-running steps to connect yt-dlp, Whisper, and ffmpeg.

The planned would use Next.js and Supabase for the website, while the video-processing worker would stay on the creator’s own computer through n8n. The local n8n worker would be exposed to the cloud backend through . The pricing idea is a free tier with a watermark and paid credit packs, with the creator paying the API costs directly.

The open questions are whether the market is too crowded because tools like OpusClip already exist, whether using n8n behind a SaaS has license risk, whether the pipeline should be rewritten in plain code before charging, whether is reliable enough for this setup, whether one computer should process jobs one at a time through a queue, and what legal issues may be missing.

Key points

  • The product idea turns long YouTube videos into short vertical clips with captions.
  • The workflow already works locally using yt-dlp, Whisper, Gemini, ffmpeg, and n8n.
  • The proposed SaaS site would use Next.js and Supabase, while processing stays on a personal computer.
  • The planned is a free watermark tier plus paid credit packs.
  • The main concerns are crowded competition, n8n licensing, local-worker , overload control, and legal risk.
Read original