One creator with 290,000 TikTok followers hasn’t paid a video editor in months. Her AI stack cuts each video in about 15 seconds, and the tools cost nothing. Here’s how to edit TikTok videos with AI the same way.
The hidden tax on every creator
Everyone wants to edit TikTok videos with AI, but editing is the part of content creation nobody warns you about. Filming takes 20 minutes. Then you sit down and cut silences, chop jump cuts, re-record the flubbed line, and export. Suddenly it’s three hours later and you still have captions to write.
Hiring an editor solves the time problem and creates a money problem. Short-form editors charge per video, and that bill lands every single week. For a creator posting daily, editing costs can quietly become the biggest line item in the whole operation.
That’s the math Brooke Lacey decided to break. She’s a creator with more than 290,000 TikTok followers and over 25 years in tech, and she explained her workflow on Lenny’s Podcast recently. She hasn’t just saved time. She rebuilt the entire weekly operation around AI and dropped her editor completely.
Her setup is worth stealing, because it runs on free tools and one subscription most readers here already have.
The free stack: Claude Code, FFmpeg, and Whisper
Three tools do all the work. None of them are actually “video editors,” which is the weird genius of this setup.
| Tool | What it is | What it does here | Cost |
|---|---|---|---|
| Claude Code | Anthropic’s terminal-based coding agent | Reads your video’s transcript and writes the edit commands | Included in Claude plans |
| FFmpeg | Open-source command-line media tool | Actually performs the cuts, trims, and export | Free |
| Whisper | OpenAI’s open-source speech recognition | Transcribes your video so Claude knows what’s in it | Free |
FFmpeg has existed since before TikTok was invented. It’s a command-line workhorse that cuts, joins, and converts media files. The catch has always been that normal humans don’t speak its weird flag language. You need a line like ffmpeg -i in.mp4 -ss 00:00:03 -to 00:00:07 -c copy clip.mp4 just to grab a few seconds.
That’s where Claude Code comes in. It speaks fluent FFmpeg so you don’t have to. You describe what you want in plain English. It writes the commands. You run them. The video gets cut.
Whisper closes the loop by turning your audio into text. Once the video is a transcript, Claude can read it, find the dead air and the rambling, and know exactly which timestamps deserve to die.
How to edit TikTok videos with AI step by step
Lacey’s live demo on the podcast showed a raw phone video getting cut in roughly 15 seconds. Your first run will take longer because of setup, but after that it’s the same pipeline every week. Here’s the shape of it:
- Get the three tools. Install Claude Code from claude.ai/code, install FFmpeg from ffmpeg.org, and grab Whisper from OpenAI’s project page. All three have beginner-friendly install docs.
- Drop your raw footage in a folder. One video per project keeps things simple. Point Whisper at the file and let it produce the transcript.
- Brief Claude on the goal. Paste the transcript in and tell it what this video is for. Lacey asks for her signature jump-cut pacing, so she says exactly that. Example instruction: “Cut every silence over half a second and tighten this for TikTok pacing.”
- Let Claude write and run the FFmpeg commands. It generates the cut list, executes it, and hands you a finished file. This is the part that took her demo about 15 seconds.
- Watch the result before posting. Check the flow, confirm no good moment got chopped, and adjust. When a cut lands wrong, tell Claude what felt off and it re-runs the edit with better timing.
The whole loop, from raw footage to ready-to-post, is maybe ten minutes of your attention. Most of that is you reviewing, not working.
Teaching the AI your style
Here’s the part most people skip, and it’s exactly why their AI edits come out generic.
The style session matters more than any single edit. Lacey didn’t just ask Claude to cut videos; she taught it her cadence, and that’s what made the output hers. It took one longer session where she showed the model examples of her edits and explained her instincts: her hooks, her rhythm, what she thinks sounds funny. Now the agent applies that style automatically on every new video.
She also has a trick for when the AI drifts. Her setup uses config files that define her editing rules, and when Claude disappoints, she prompts it back on track and updates the file so the same mistake doesn’t happen twice. The agent gets smarter about her style every week instead of resetting to zero.
That’s the real difference between “AI edited my video” and “AI edits my videos.” One is a one-off trick. The other is a system that compounds.
Your comments are the content brief
The second half of her workflow has nothing to do with cutting footage, and honestly it’s the more valuable half.
Lacey mines her TikTok comments for scripts. When people ask questions or argue in the replies, those become the next videos. So the audience is literally writing her content plan, and she shows up once a week to film and let the AI pipeline handle the rest.
As a result, one focused filming session produces her entire week of posts. She grows the audience while posting far less than the daily grind most creators force on themselves. Meanwhile the comment mining keeps every video pre-validated: she knows people want the answer before she films it.
If you’re juggling creator work with a day job, this pairing is the whole game. Batch your filming, automate your editing, and let the comments decide the topics. Creators who want the full toolkit should also look at what Apple Creator Studio AI features include, since that covers another slice of the production workflow for cheap.
What this AI editing workflow can’t do
Fair warning about what breaks when you edit TikTok videos with AI every week, because the demo reels never show the misses.
- B-roll and visuals. This pipeline cuts talking-head footage. It doesn’t film inserts, grab memes, or design thumbnails. You still do the creative garnish.
- Taste on first pass. The AI cuts by rules and learned style. If your pacing instinct is unusual, expect a few rounds of correction before it matches your brain.
- Zero setup magic. The 15-second edit comes after tools are installed and the style session is done. Budget an afternoon for the first run. After that it pays you back every week.
One more thing worth saying. This is the same one-person-with-agents pattern we saw in the Devin case study, where a solo founder ran a whole company’s output through AI workers. The tools change. The advantage doesn’t. And if you’d rather generate whole videos instead of editing real footage, our Gemini Omni guide covers Google’s AI video maker, which handles a different kind of content job.
The takeaway
You can edit TikTok videos with AI for the cost of a Claude subscription, and the workflow above is already proven at 290k-follower scale. Set up the three tools this week, run one old video through the pipeline, and see if the result passes your own scroll test. If it does, you just fired your most expensive weekly task.
Sources: Brooke Lacey on Lenny’s Podcast, Claude Code, and FFmpeg.org.