S
Claude CodeFaceless long-form videos with zero on-camera time
Faceless Channel: Script to Upload in One Sitting
setuproll@setuproll91.0Overall score
A full assembly line that turns a single topic into a finished, narrated, captioned YouTube video without you ever facing a camera. Built for solo creators running history, finance, or explainer channels who want to publish two or three videos a week.
91.0Score
2.8kVotes
4Components
Install this build
terminal
npx setuproll init youtube-faceless-factoryComponents
Model
- Claude Opus 4.8 (outline + script)
- Claude Sonnet 4.6 (titles + chapters)
Stack
- ElevenLabs (voiceover)
- Runway Gen-4 (b-roll)
- CapCut (assembly + captions)
Publishing
- vidIQ (titles + tags)
- Canva (thumbnail)
- YouTube Studio
How it works
- Feed the topic to Opus 4.8 with a retention-focused outline prompt, then expand to a spoken-word script with hooks every 30 seconds
- Send the script to ElevenLabs for a consistent narrator voice, exporting one MP3 per chapter
- Generate or pull b-roll in Runway Gen-4 and stock libraries, matching each script beat
- Drop audio and clips into CapCut, auto-caption, then export 1080p or 4K
- Use vidIQ to lock the title, description, and tags before uploading via YouTube Studio
Summary
A full assembly line that turns a single topic into a finished, narrated, captioned YouTube video without you ever facing a camera. Built for solo creators running history, finance, or explainer channels who want to publish two or three videos a week.
91.0 score 2.8k votes
0 Reviews
Your rating
Sign in to post
Loading discussion...