How to Make YouTube Shorts with AI (Full Guide)
Shows filmmakers how reverse-engineering proven video structures with AI scripting and targeted image models enables rapid, repeatable short-form video production without technical overhead.
What this lesson covers
Youri van Hofwegen outlines a 5-step workflow to quickly produce viral YouTube Shorts by leveraging existing viral structures. He demonstrates using Claude for script and prompt generation, ElevenLabs and GPT Image 2 via OpenArt for assets, and CapCut for final timeline editing.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 01:15 ↗
Extracting script transcripts from high-performing videos allows creators to replicate successful content structures without copying original text.
- 01:42 ↗
Claude can analyze competitor transcripts to reverse-engineer viral pacing and re-write niche-specific scripts matching the original tone.
- 02:57 ↗
OpenArt's ElevenLabs integration with the 11v3 model enables voiceover filtering by accent, gender, and conversational style.
- 04:25 ↗
Explicitly specifying the target image generator model in Claude ensures prompt structures align with model preferences.
- 05:25 ↗
Setting image resolution to 4K and aspect ratio to 1:1 in GPT Image 2 produces clean meme-style assets for short-form video edits.
Workflow outlined in the video
- Find a viral YouTube video in a adjacent niche and extract its transcript using an online transcript scraper.
- Input the transcript into Claude with a prompt to write a new script for your specific topic using the original structure.
- Ask Claude to generate scene-by-scene prompts tailored specifically for GPT Image 2.
- Generate voiceovers using the 11v3 model in OpenArt, selecting a conversational male or female voice.
- Render 1:1 aspect ratio, high-resolution visual assets with GPT Image 2 in OpenArt.
- Assemble the voiceover track, AI images, and non-copyrighted gameplay footage inside CapCut.
Before you use this workflow
These notes describe the source video at its publication date. Model access, pricing, connectors and interfaces may have changed. Check the original source and the provider’s current documentation before installing an add-on, connecting an account or spending credits.
ReelStack has not independently tested this workflow. Preview one representative shot and check motion, continuity and output quality before applying it to a full production. No result, cost saving or model capability is guaranteed.
Explore the tools
How to interpret the numbers
331,658 views captured 25 October 2026. ReelStack recommendation score: 56/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 45% performance: 5.59× views vs 11 other channel videos observed at a similar age. View-evidence factor 100% (views / (views + 1,000))
- 25% freshness: 0/100 with a 14-day half-life
- 20% momentum: 4236 views/day vs 403 channel baseline (5 comparable uploads), with the same view-evidence factor
- 10% engagement: 1/100; likes + 4× comments, smoothed with a 500-view neutral prior