This is the operational guide. Not theory. Not vague advice. The actual step-by-step process that creators use to turn 45 minutes of long-form content into 10 polished shorts in under an hour. Start with the complete pillar guide on AI short-form video if you want the full context.
We're using Opus Clip as the primary tool, with notes on how to adapt this for Munch or Vizard. This workflow assumes you have one piece of long-form content: a podcast episode, YouTube video, Twitch stream recording, or interview.
Total time: 40 minutes from raw video to 10 scheduled shorts. Processing time: 10 minutes. Your active work time: 30 minutes. That's 3 minutes of work per short.
The Complete Workflow: Day by Day
Monday: Record or Source Your Long-Form Content
Record a 45-50 minute piece of content. It can be a podcast episode, YouTube video, Twitch stream, interview, educational lecture, vlog, or anything else. Quality matters here — good audio, reasonable video quality, stable framing. Don't obsess, but don't phone it in.
Where to save it: Upload it to your computer or cloud storage. File format doesn't matter (MP4, MOV, whatever). You'll upload it to the clipping tool next.
Pro tip: If you're recording content specifically for repurposing into shorts, film it in a way that supports clipping. Natural speaking breaks. B-roll changes. Visual transitions. These make AI clipping significantly easier.
Tuesday Morning: Upload and Let It Process
Step 1: Sign up for Opus Clip (if not already)
Go to opus-clip.com. Sign up (Google or email). Create an account. Done. Free tier available.
Step 2: Upload your video
Click "Create Clips". Select your video file. For free tier, max 60 minutes. If over 60 minutes, trim it first or upgrade. Hit upload. Opus Clip immediately starts processing — you can close the browser and walk away.
Step 3: Wait for processing
Opus Clip analyzes your video. Transcribes it. Identifies natural speaking breaks and moments. Generates 10-15 potential clips. Processing time depends on video length and your internet speed. Go grab coffee.
Tuesday Mid-Morning: Review and Delete Duds
Step 4: Review generated clips
Processing finished? Check your Opus Clip dashboard. You now have 10-15 clips. Watch each one (most are 20-60 seconds, takes 2 minutes total to scan them). Mentally sort them into three buckets: "Gold" (definitely keeping), "Maybe" (good but needs tweaking), "Trash" (delete).
Step 5: Delete the trash
Delete clips that don't work. Bad audio. Random breakpoint. Boring moment. Don't overthink — if you're not sure, keep it. You can always delete later. Target: 8-10 good clips after deletion.
Tuesday Afternoon: Customize and Add Captions
Step 6: Tweak individual clips (optional)
For each "Maybe" clip: adjust the in-point and out-point if needed. Opus Clip lets you drag to adjust timing. If the first second doesn't grab attention, trim from the start. If the ending feels weak, trim the end. Most clips need zero tweaks. Some need 10 seconds of adjustment.
Step 7: Add captions with Submagic
Opus Clip adds basic captions. If you want animated captions (highly recommended for TikTok and Reels), use Submagic. Download each clip from Opus Clip. Upload to Submagic. Generate animated captions. Edit any errors. Download back. Takes 60-90 seconds per clip. For 10 clips: 10 minutes total.
Submagic quick start: Upload video → "Generate Captions" → Edit any wrong words → Choose animation style → Download. That's it.
Wednesday Morning: Add Hooks and Refine
Step 8: Generate hook variations with ChatGPT
For your strongest clips, generate hook variations. Open ChatGPT. Use this prompt template:
"I'm creating TikTok videos about [topic]. The clip is [brief description of what happens in the clip]. Generate 20 hook variations. Hooks should be conversational, grab attention in the first 2 seconds, and relate specifically to what I'm saying in the video. Don't be clickbait — be honest."
ChatGPT generates 20 variations in 30 seconds. Pick your five favorites. Edit them to match your voice better. You now have 5 on-screen text hooks ready to add.
Step 9: Add hooks as on-screen text
Use CapCut or Submagic to add your best hook as on-screen text at the start of your clip. Or manually add it in your editing tool of choice. Position it prominently. Make it readable. 30-60 seconds of work per clip.
Wednesday Afternoon: Audio and Final Polish
Step 10: Match audio to clip (optional but recommended)
Does your clip match trending TikTok audio? Use Suno AI to generate custom music, or manually add royalty-free music from Epidemic Sound or YouTube's library. Trending audio can 5-10x your views if it matches. Non-trending audio still works. No audio feels flat.
Quick check: If your video has natural audio (voice, conversation, reactions), keep it. That's usually better than added music. But if there's silence or background noise, layer in music.
Step 11: Final check and export
Watch each clip one more time. Does the hook grab you? Does the audio make sense? Is the caption readable? Export as MP4 in vertical format (9:16 aspect ratio for TikTok and Reels, 1:1 for Stories).
Thursday: Schedule and Publish
Step 12: Schedule on TikTok, Reels, Shorts
Use a scheduler like Buffer, Later, or post directly to each platform. Don't post all 10 on the same day — space them out. Ideal schedule: 2-3 per week over 4 weeks. Or 1 per day for 10 days. Stagger to avoid algorithm suppression and maintain consistent presence.
For each platform:
- TikTok: Post natively (no external scheduler) for best algorithm treatment. Or use TikTok's built-in scheduler.
- Reels: Post natively to Instagram, or use Meta's scheduler (part of Business Suite).
- Shorts: Post natively to YouTube. YouTube has a built-in scheduler.
Pro tip: TikTok's algorithm gives a small boost to native posts versus cross-posted. Same with Reels and YouTube. If you're serious, spend an extra 3 minutes per platform to post natively instead of using a cross-poster.
Timeline Summary
| Day | Task | Time |
| Monday | Record long-form content | 45 min (your own time) |
| Tuesday | Upload, process, review, delete duds | 15 min active |
| Wednesday | Captions, hooks, audio, polish | 20 min active |
| Thursday | Schedule and publish | 10 min active |
Variations: Using Munch or Vizard Instead
If You're Using Munch
Munch's workflow is similar but with automatic platform optimization. After step 3 (processing): You get clips already optimized for TikTok vs Reels vs Shorts. Instead of manually adjusting for each platform, Munch handles it. Skip step 6 (manual tweaking) and 10 (audio) because Munch gives you options for those. Total time: still 40 minutes, but more of it is clicking options instead of manual adjustment.
If You're Using Vizard
Vizard is free with a watermark. Workflow is nearly identical to Opus Clip. Processing might be slightly slower. Output quality slightly lower. Steps 4-5 take longer because more clips need deletion. But process is the same. Time: 40-50 minutes total.
Common Problems and Solutions
Problem: I got clips, but many are bad.
Solution: This means your source video isn't ideal for clipping. Either the audio quality is poor (AI can't find natural breaks), or the pacing is slow (every moment feels like a break), or the visuals are repetitive (no visual interest to anchor clips). Next time: record with clipping in mind. Natural speaking pauses. Visual changes. B-roll transitions.
Problem: Processing is taking too long.
Solution: Longer videos take longer. If you're uploading 90 minutes, it'll take 15+ minutes. Trim your source video to 45-60 minutes before uploading. Also: internet speed matters. If you're on slow internet, processing takes longer because the upload takes longer.
Problem: I don't have time to customize each clip.
Solution: Skip steps 6-10. Post the AI-generated clips as-is. You'll lose maybe 10-20% of potential performance, but you'll save 20 minutes. As you see which clips perform, come back and customize the winners. Build up the process iteratively.
Problem: My captions are wrong.
Solution: AI captions are never perfect. Always review and edit. Submagic and Descript both let you manually edit captions. Spend 30 seconds per clip fixing errors. It's worth it — correct captions perform better and are more professional.
Next Level: Scaling This Workflow
Once this workflow is comfortable (week 2-3), optimize it further:
- Batch processing: Record 2-3 hours of long-form content on recording day. Process all of it. Generate 20-30 shorts. Schedule one per day for the month. You recorded once, generated content for 30 days.
- Automation tools: Use Repurpose.io to partially automate the workflow. It's not perfect, but it handles 40-50% of the process.
- Template hooks: Create 10-20 proven hook templates and reuse them. ChatGPT generates variations, you pick from templates. Faster than custom every time.
- Audio library: Build a playlist of trending audio on TikTok that matches your content type. Match clips to pre-existing audio instead of generating new music every time.
The Math: Why This Works
Manual editing: 45 minutes of content = 6-8 hours of editing to generate 10 shorts. With AI: 40 minutes total. That's 16x faster. At scale: if you publish 40 shorts per month, manual takes 240 hours. AI takes 160 minutes (2.5 hours). The difference is 237.5 hours per month — that's nearly 6 full workweeks saved.
That's why creators with AI are winning. They're not smarter or more creative. They're just faster. And in short-form, speed is strategy. More tests, more variations, more audience feedback, better content.
Start this workflow Monday. By Thursday you have 10 scheduled shorts. By next Monday you know what worked. Iterate. That's how you grow on TikTok.
