The Zella AI Assistant: Edit Your Video by Typing

Quick answer: The AI Assistant is a chat box inside Zella’s editor. You type what you want in normal English — like “remove the quiet parts, add captions, and make it tall for TikTok” — and it does the editing for you. To turn it on, open a project, click the Assistant tab, and add a free AI key (takes about a minute). Every change can be undone by pressing ⌘Z.

This page is a full menu of what you can say. You do not need to memorize anything. Just describe what you want and the assistant figures out the rest. Below, every command is written as “You type” and “What happens” so you can copy the examples and change the words to fit your video.

Jump to a section: What it is · Set it up · How to talk to it · Cutting · Cleaning up · Captions · Audio · Color & looks · Background · Zoom · Speed · Text & callouts · Overlays · Resize · Content search · Seeing the frame · Undo · Export · FAQ

What is the AI Assistant?

Think of it as a helper that sits next to your video and knows how to use every button in Zella. Instead of you clicking around to find the right tool, you tell it what you want and it clicks for you.

It can do over 70 different editing jobs — cutting, adding words on screen, fixing your voice, changing colors, adding music, resizing for social media, and more. This page walks through all of them, grouped by what you are trying to do.

You can give it one job at a time, or several jobs in one sentence. Both work.

Setting it up (one time)

The assistant needs a “key” to connect to an AI brain that understands your words. Several good options are free and take a minute to set up.

Open any project in Zella and click the Assistant tab on the right.
Click the gear icon in the top corner of the chat.
Pick a provider. For a free choice with no credit card, choose Groq, Google Gemini, or Cerebras. (If you want it to run fully on your own Mac with no internet, choose Ollama.)
Click Get free key. That opens the provider’s website. Sign up, copy the key they give you, and paste it into Zella.
Click Save Key, then Test Connection. When it says connected, click Done — start chatting.

Is this safe? Yes. Your key is stored in your Mac’s private Keychain. Zella never sends your video anywhere. The only thing that leaves your Mac is the sentence you type plus a short summary of your project (like how long it is), and that goes only to the provider you picked — the same as typing into their own app.

How to talk to it

Type the way you would ask a friend. You do not need special words. All of these work:

“Remove the silences.”
“Can you take out the boring quiet bits?”
“Cut the dead air, please.”

If you are not sure what to say, type “what can you do?” and it will list its main skills. If it needs to know which clip you mean, it will look at your project and figure it out (clips are numbered starting at clip 1).

For big actions that cannot be easily reversed (like deleting all silences, or exporting), it asks you to confirm first. Everything else you can simply undo with ⌘Z.

Now here is everything you can ask for.

1. Cutting and trimming

Split the video where the playhead is

You type: “Split here” or “Cut at the playhead.”
What happens: It adds a cut line at the current spot, breaking the video into two pieces so you can treat them separately.

Split at a specific time

You type: “Add a cut at 15 seconds.”
What happens: It places a cut exactly at 0:15.

Delete a piece by its number

You type: “Delete clip 2.”
What happens: It removes the second piece and slides everything after it back to close the gap. (This is called a “ripple delete” — the timeline ripples closed so there is no empty hole.)

Delete a range of time

You type: “Cut out everything from 10 to 15 seconds.”
What happens: It deletes that 5-second stretch and closes the gap.

Reorder pieces

You type: “Move clip 3 to the beginning.”
What happens: It rearranges the order of your cut pieces.

Set the start and end (trim)

You type: “Trim it to start at 5 seconds and end at 40.”
What happens: It sets where the finished video begins and ends, hiding anything outside that range.

2. Cleaning up the boring parts

Remove silent gaps

You type: “Remove all the silent gaps.”
What happens: It finds every quiet pause where nobody is talking and deletes them, so your video feels tighter. It shows you how many it found and asks before applying.

Remove filler words

You type: “Cut out the ums and uhs.”
What happens: It listens for filler words like “um,” “uh,” and “like” and removes them so you sound more confident.

One-click polish (does several at once)

You type: “Magic edit” or “Clean this up for me.”
What happens: It removes background noise, smooths your voice, brightens the colors, and adds captions — all in one go. Nothing gets cut, so it is safe to try.

3. Captions (words on screen)

Add captions

You type: “Add captions.”
What happens: It listens to your audio and types out what you said as on-screen captions. This all happens on your Mac.

Change the caption style

You type: “Use the Word Pop caption style” (10 styles: Word Pop, Hormozi, Karaoke, Pop Box, Neon Glow, Clean Minimal, Beast, Highlighter, Bounce, Thumbnail).
What happens: It restyles all your captions to the look you named.

Edit one caption line

You type: “Fix caption 3 to say ‘Welcome back’.”
What happens: It changes the words, timing, or highlighted word of a single caption line.

4. Audio (voice, music, and sound effects)

Set the volume

You type: “Make clip 2 quieter” or “Set the volume to 50%.”
What happens: It lowers or raises the loudness, either for one clip or the whole video.

Add background music

You type: “Add background music from /Users/me/Music/song.mp3.”
What happens: It drops your music file under the video and automatically keeps it quiet enough to hear your voice. (You give it the file location; music is not built in.)

Add a sound effect

You type: “Add a whoosh at 3 seconds” or “Put a vine boom here.”
What happens: It adds one of 30 built-in sound effects (whoosh, pop, riser, air horn, ta-da, camera shutter, and more). No file needed — these come with Zella.

Clean up your voice

You type: “Clean up my voice” or “Remove the background noise.”
What happens: It removes hiss and hum and evens out your voice so it sounds clear and professional.

Duck the music under your voice

You type: “Lower the music when I talk.”
What happens: It automatically dips the music volume whenever you are speaking, then brings it back up.

Change your voice tone

You type: “Make my voice warmer” (or brighter, podcast, cinematic).
What happens: It adjusts the tone color of your voice.

Move audio earlier or later

You type: “Make the audio start a little before the video.”
What happens: It shifts the sound slightly ahead of or behind the picture (useful for a smoother, more natural cut).

5. Color and looks

Auto-enhance

You type: “Auto enhance the colors.”
What happens: It gently improves brightness, contrast, and color in one step.

Apply a look

You type: “Make it look cinematic” or “Give it a Film Noir look.”
What happens: It applies a ready-made color style (like Teal & Orange, Kodachrome, or Film Noir).

Apply a look to every clip

You type: “Make every clip cinematic.”
What happens: It puts the same color style on all your clips at once.

Adjust one setting

You type: “Make it brighter” or “Turn up the contrast.”
What happens: It changes a single color control like brightness, contrast, saturation, or warmth.

Add a special effect

You type: “Make it black and white” or “Add a vignette.”
What happens: It applies effects like black-and-white, blur, vignette (dark edges), grain, or glow.

Start over on colors

You type: “Reset the effects on this clip.”
What happens: It removes all color changes and effects from the selected clip.

6. Background

Remove your background

You type: “Remove my background.”
What happens: It cuts you out from behind and blurs whatever was there (great for webcam videos). This runs on your Mac and can take a little while.

Replace the background

You type: “Put me on a navy background” or “Use a solid blue background.”
What happens: It places you on a new background — a built-in scene, a solid color, or your own image or video.

Undo the background change

You type: “Bring back my original background.”
What happens: It puts your real background back.

7. Zoom

Zoom in on a spot

You type: “Zoom in here.”
What happens: It adds a “punch-in” zoom at the current spot to draw attention.

Auto-zoom on your clicks

You type: “Add zooms where I clicked.”
What happens: If you recorded your screen, it automatically zooms in on the places where you clicked your mouse (like Screen Studio does).

Adjust a zoom

You type: “Make the first zoom stronger” or “Delete zoom 2.”
What happens: It changes how far a zoom goes, how long it holds, or removes it.

8. Speed (fast and slow motion)

Speed up or slow down a clip

You type: “Speed up clip 2 by 2x” or “Make clip 1 slow motion.”
What happens: It changes the playback speed. 2x means twice as fast; 0.5x means half speed (slow motion).

Add a smooth speed change

You type: “Add a speed ramp to clip 3.”
What happens: It smoothly speeds up or slows down instead of changing all at once.

9. Titles and callouts

Add a title (big text)

You type: “Add a title that says Welcome.”
What happens: It puts a text title on the video at the current spot.

Add a callout (a pointer)

You type: “Add an arrow” or “Put a spotlight here.”
What happens: It adds a shape that points at something — arrow, box, circle, label, keystroke, spotlight, number, or blur.

Edit a callout

You type: “Move callout 1 to the top and make it say ‘Click here’.”
What happens: It changes a callout’s position, text, or timing.

10. Overlays (logos, images, and picture-in-picture)

Add an image or logo

You type: “Add my logo from /Users/me/Pictures/logo.png at 2 seconds.”
What happens: It places the image on top of your video as a layer you can move and time.

Add a second video (picture-in-picture)

You type: “Add the b-roll clip from /Users/me/Movies/broll.mp4.”
What happens: It lays a second video over your main one (like a small video in the corner), with its sound mixed in.

Edit a picture-in-picture video

You type: “Make video overlay 1 quieter” or “Move it to start at 5 seconds.”
What happens: It changes the overlay’s timing, length, or volume.

11. Resize for TikTok, YouTube, Instagram

Change the shape for a platform

You type: “Make it vertical for TikTok” or “Set it to YouTube.”
What happens: It changes the video’s shape (tall 9:16 for TikTok/Reels/Shorts, wide 16:9 for YouTube, square 1:1 for feeds).

Fill or fit

You type: “Fill the frame” or “Fit it with black bars.”
What happens: “Fill” crops the video to fill the whole screen; “fit” shrinks it so nothing is cut off, adding bars if needed.

12. Find parts by what you said

The assistant can read the words you spoke, so you can edit by content instead of hunting through the timeline. (If it has not listened yet, just say “add captions” first.)

Jump to something you said

You type: “Go to where I say welcome.”
What happens: It searches your speech and moves the playhead there.

Cut something you said

You type: “Cut the part where I fumble the intro” or “Delete the sentence about the old price.”
What happens: It finds those words in your speech, then removes that stretch of video.

13. It can look at your video

Ask the assistant about the picture and it will actually check the frame on your Mac (no upload).

Check your shot

You type: “Is my face centered?” or “Is this too dark?”
What happens: It looks at the current frame and tells you what it sees — where your face is, how bright it is, what is in view — and can fix it if you ask.

14. Undo and redo

Undo

You type: “Undo that” — or just press ⌘Z.
What happens: It reverses the last change. You can undo up to 50 steps.

Redo

You type: “Redo” — or press ⇧⌘Z.
What happens: It brings back the change you just undid.

15. Export (save the finished video)

Export

You type: “Export the video” or “Export in 4K for TikTok.”
What happens: It saves your finished video to your Downloads folder. You can name the quality (720p, 1080p, 4K, 8K) and the platform. It always asks you to confirm first because it creates a file. (4K and 8K need Zella Pro; 1080p is free.)

Rename the project

You type: “Rename this project to Launch Demo.”
What happens: It changes the name shown in your Library.

FAQ

Do I have to remember these commands? No. Just describe what you want in your own words. This page is only a menu of ideas.

Is it free? The assistant is part of Zella Pro, but it works with free AI providers (Groq, Gemini, Cerebras, or Ollama on your Mac), so there is no cost per edit. You bring your own key.

Does my video get uploaded? No. Zella never uploads your video. Only your typed message and a short text summary of the project go to the AI provider you chose. Pick Ollama to keep everything on your Mac with no internet.

What if it does something I do not like? Press ⌘Z to undo. You can also read each step it took in the chat before you export.

Can it do everything the buttons do? Almost. It covers the large majority of edits. For tiny, exact tweaks (like nudging a title a few pixels) the panels are still handy — and you can freely mix typing with clicking, since they share the same timeline.

Which provider should a beginner pick? Start with Groq — it is free and fast. If you ever want zero internet, install Ollama and everything runs on your Mac.