Editing a talking video by hand follows the same routine every time: cut the silences and false starts, add captions, drop in some B-roll, place a title, add sound effects and music, balance the audio, export. It is skilled work, and it is also repetitive. An AI video editor does that routine for you, in the style you choose, and leaves you a normal timeline you can change.
This guide explains how the VideoLab AI Editor works, how to prepare your footage, how to pick a style, how to write a description that steers the edit, and how to finish the video by hand.
What the AI Editor does
You give it raw footage and a style. It builds a complete edit on a real timeline:
- Cuts the silences and dead air between sentences, so the video moves.
- Adds captions word by word, in the look of the style.
- Places B-roll, photos, titles and cards at the right moments, using the files you provide.
- Adds sound effects and music, mixed under your voice.
- Applies the style's pacing, scenes and text looks.
The edit is not a black box. You watch it being built: a cursor takes your video, places it on the timeline, listens to what is said, jumps to each silence and cuts it, then lays down captions, images, titles and sound one by one. When it is done, everything is a normal clip on the timeline that you can move, trim or delete. The whole automatic edit is a single undo step, so one Ctrl+Z takes you back to your raw footage.
Step 1: Your video
Start a new AI Editor project and add:
- Your footage (required). One or several videos. If you add several, they are used in the order you give them.
- Images and B-roll (optional). Photos become pop-ups and illustrations; extra video clips become B-roll. The more relevant material you give, the richer the edit.
- A title (optional). It names the project.
- A format. Auto picks the format closest to your first video, or choose 9:16 (TikTok, Reels, Shorts), 1:1, 4:5 (Instagram feed) or 16:9 (YouTube).
- The spoken language. Auto-detect, Mixed languages, or a specific language.
- Captions. Auto, Always or Off.
Step 2: Choose a visual style
The style is the biggest decision. Each style in the VideoLab catalog defines the cuts, the rhythm, the scenes and the look of the captions together, so the whole video feels consistent. You can preview each one: a short video plays on hover, and the preview tells you what the style is best for and what to avoid.
Some styles have extra fields, for example a title, a price or a list of items, shown under the card when you pick it. Fill them in: they become the on-screen text of the edit.
How to choose:
- Talking head, advice, education: a clean style with captions and occasional B-roll.
- Product, real estate, before/after: a format with photo grids, prices and titles.
- Story or vlog: a style with scenes and music.
- Quote or statement: a quote format that puts your strongest sentence on screen.
You can also save your own styles and reuse them on every video, so your channel keeps one recognisable look.
Step 3: Describe what you want (optional)
The Describe step lets you tell the AI, in plain words, what this video is and how you want it edited. You can skip it, but a short description makes the edit fit your intent better. Useful things to say:
- The goal: "A 45-second tip video that ends with a call to follow."
- The audience: "For first-time home buyers."
- The tone: "Calm and premium, no flashy effects." or "Fast and funny."
- What to keep or cut: "Keep the joke at the start. Cut the part about shipping."
- What to emphasise: "Put the three numbers on screen as titles."
The app suggests ideas you can add with one click.
Prepare your footage for a better edit
The AI works with what you give it. Five minutes of preparation saves twenty minutes of fixing:
- Clear audio first. The edit is built on the transcript: clean speech means accurate cuts and captions. Use a microphone close to your mouth and avoid echoing rooms.
- Pause between ideas. Short pauses give the AI clean places to cut.
- Restart a sentence instead of fixing it mid-way. Say it again from the beginning; the silence before the retake helps the cut.
- Provide B-roll. Even five or six short clips or photos make the edit far richer than a talking head alone.
- Film in the format you will publish. Vertical for shorts, horizontal for YouTube, so the AI does not have to crop your frame.
Read more in how to film for AI editing.
What about videos with little speech?
If most of your video has no clear speech, like a renovation, a landscape or a cooking video without voice-over, the AI Editor does not cut it as a talking video. It keeps the footage whole, does not remove the "silences", and captions only the words it is sure about. You can then add music, titles and B-roll on top.
Finish the edit by hand
The AI gives you a strong first cut. The finishing touches are yours, and editing by hand is always free:
- Watch the first five seconds. Is the hook clear? Add a title if not. The AI packaging tool suggests hook lines from what you say.
- Check the cuts. If a cut feels abrupt, drag the edge of the clip to give it a few more frames.
- Correct the captions. Click any caption to fix a word.
- Swap B-roll that doesn't match. Drag another file from your media onto the B-roll track.
- Balance the sound. Music should never fight your voice. Lower it or use the sound check.
- Run the pre-export check. It flags black gaps, sound or text after the last image, leftover tiny clips, missing files, text under the TikTok/Reels buttons and captions placed too low.
Export
Choose a file name and where to save, then export. You can export several formats at once, for example the same video in 9:16 and 16:9, and VideoLab also exports GIF, MP3 and project files for Premiere Pro and DaVinci Resolve if you want to finish elsewhere.
- Free: exports up to 1080p. Videos edited by the AI carry a watermark on Free.
- Lab+: no watermark and exports up to 4K.
AI Editor or Clip Factory?
Both use AI, but for different jobs:
| You have… | You want… | Use |
|---|---|---|
| Raw footage of one video | One finished, edited video | AI Editor |
| A long video, podcast or stream | Several short clips from it | AI Clip Factory |
If you add a long video to the AI Editor, the app reminds you that the Clip Factory is the right tool for getting short clips out of it.
Frequently asked questions
Will an AI video editor replace my editor?
It replaces the repetitive part of the job: silence cuts, captions, first placement of B-roll and sound. The creative choices, the story and the final polish still benefit from a human eye. Many editors use it to deliver faster.
Can I change the edit after the AI is done?
Yes. Everything lands on a normal timeline: every cut, caption, image and sound can be moved, changed or deleted. One undo returns to your raw footage.
How long does the automatic edit take?
It depends on the length of your footage and on your device. You can watch the edit being built live, or press "Skip to result" to jump to the finished timeline.
How many AI minutes does an edit use?
1 minute of footage sent to the AI equals 1 minute of credit. A 3-minute video uses about 3 minutes. Free accounts get 30 minutes a month; Lab+ includes 600.
Does it work for languages other than English?
Yes. Choose the spoken language or let the app detect it; captions follow the language of the video.