HOW IT WORKS
Record. One tap. Posted.
Two questions decide whether a video tool is worth your time: will it actually work on my footage, and how much work is it? This page answers both with a clock on it. The short version - one tap after you stop recording produces a finished video, and finding that out costs nothing.
Quick answer
Making a video in Zella takes three moves: record, tap Auto-Polish, export. For a 60-second talking-head short that is about 12 minutes end to end - roughly 5 minutes recording, under a minute for Auto-Polish, 3 to 4 minutes reviewing, and about a minute to export. Zella runs entirely on your Mac or iPhone, and finding out costs nothing: the free tier exports up to 2 minutes at 1080p with no watermark and no card.
THE SHORTEST PATH
Three moves. Two of them are yours.
This is the entire route, not a simplified version of it. Someone who does only this ends up with a cut, captioned, voice-cleaned and scored video, in the colour they shot it in. Everything else on this page is optional depth.
Record
YOU DO
Pick a source and talk. Fumble a line, pause, say it again, keep rolling - the edit fixes it later.
ZELLA DOES
Captures screen and camera with system and mic audio, segments the take so a crash, a call or a flat battery cannot cost you the recording, and opens the editor the moment you stop.
HOW LONG
As long as your take. About 5 minutes of talking for a 60-second short, if you retry in-roll instead of restarting.
Tap Auto-Polish
YOU DO
One tap. Then watch it back.
ZELLA DOES
Runs the full retention edit: silence and filler cuts, a cut cadence with varied transitions, a matched sound effect under every cut, keyword captions, noise removal and voice polish, a hook title with an opening punch-in, emphasis zooms, mood-matched ducking music and pattern interrupts. Every step is undoable.
HOW LONG
Typically under a minute on a short take. Add 3 to 4 minutes if you watch it through and nudge a few things, which most people do once and then stop bothering.
Export
YOU DO
Pick the platform. Hit export.
ZELLA DOES
Reframes to 9:16, 1:1 or 16:9 with face-aware tracking, applies the right one of the platform presets, and renders on your own machine. No queue, no upload, no watermark on any tier.
HOW LONG
About a minute for a 60-second 1080p clip, then it is a file on your device.
About 12 minutes, opening the app to having a file. That is the real number for a 60-second talking-head short - roughly 5 minutes recording, under a minute of Auto-Polish, 3 to 4 minutes watching it back, a minute to export. Nothing is queued on a server, so none of it is spent waiting on someone else's machine.
THE ONE TAP, UNPACKED
Eleven passes, in the order that matters.
"One tap" is only worth anything if you know what the tap does. Auto-Polish runs these in sequence, each one built against the timeline the last one left behind - which is why the cuts come before the captions and the pattern interrupts come last.
- 01
Cuts the silences
Dead air between sentences is ripple-deleted, with padding so trailing consonants finish. A rambling take typically tightens by 20 to 40%.
- 02
Cuts the filler words
Um, uh and like come out of the same pass, read from the on-device transcript.
- 03
Sets a cut cadence
It checks the whole take and adds cuts at natural pauses so no stretch of frame sits unchanged for long, targeting roughly five-second intervals.
- 04
Varies the transitions
Every join gets its own transition rather than the same dissolve twenty times.
- 05
Sounds the cuts
A matched effect sits under each transition at about half voice volume - impacts on the cut, whooshes straddling it.
- 06
Captions it
On-device speech recognition, then big keyword-emphasised captions. Nothing is sent anywhere to be transcribed.
- 07
Cleans the voice
Neural noise removal, de-ess, compression and loudness normalisation, so a kitchen table sounds like a desk.
- 08
Builds the opening
A hook title from your first strong line, paired with a punch-in that starts close and settles out over roughly 1.7 seconds.
- 09
Adds emphasis zooms
Punch-ins land on the moments that carry weight, so a static frame stops being static.
- 10
Scores it and ducks it
A music bed mood-matched to your pace, snapped to the first beat, swelling in pauses - and Auto-Duck rides it under every word.
- 11
Fills the flat spots
A final pass scans the finished edit for any window still visually static and breaks it up, without doubling on moments the earlier passes already animated.
On colour, Auto-Polish does nothing at all - deliberately, on every device. A look you already like survives the tap, and footage you have not touched comes out the colour you shot it in. Grading stays a separate control you reach for on purpose: 43 looks on the Mac and 42 on iPhone with an intensity slider and your own .cube LUTs, free on every tier.
Record
On the Mac, click the menu-bar icon and pick full screen, a region, a window, or camera-only. Add a webcam bubble, choose a clean background, and capture system plus mic audio. On iPhone, tap record for the camera, the screen or the whole device, with a background you can blur, cut out or replace live while you film. Pause and resume anytime - long takes survive sleep, calls and crashes.
Edit
The timeline opens with your take already on it. Auto-Polish is the one tap that finishes the video. Prefer to steer? Type it: the AI Assistant takes plain English and does the editing, and it is free on both platforms. Prefer your own hands? The timeline is a real one - B-key cuts, ripple delete, drag-to-reorder, J and L cuts, 50-level undo.
Ship
Reframe to 9:16, 1:1, or 16:9 with face-aware auto-track, then export MP4, MOV, GIF, or M4A. Zella recommends the right platform preset - TikTok, Reels, Shorts, YouTube, Instagram - one click away. Free exports are watermark-free at 1080p, up to two minutes. Pro lifts the length cap and adds 4K, plus 8K on the Mac.
TWO WAYS IN, ONE EDITOR
Start wherever the video is.
The Mac app and the iPhone app are the same studio with different hands on it. Whichever one you start in, the recording lands in the same editor.
ON THE MAC
It lives in the menu bar.
Zella is not a window you go and open. Click the menu-bar icon, pick your sources and a region, and press Start Recording. When you stop, Studio opens by itself with the take already on the timeline. Two windows total: the setup panel you record from, and the editor you finish in.
- Menu-bar icon to recording in about three clicks
- Studio opens automatically the moment you stop
- Full keyboard set, including a one-key cut
- macOS 14 Sonoma or later, Apple Silicon and Intel
ON IPHONE & IPAD
Open the app and record.
Front or back camera, your screen, or a whole-device broadcast - with a background you can blur, cut out or replace while you film. Tap stop and the editor opens with your take ready to cut. The whole loop runs on the device, so it works on a plane and a client property never leaves your phone.
- Camera, screen or whole-device capture
- Live background blur, cutout or replace at capture
- Segmented takes, so a phone call cannot cost you the recording
- Zella for iPhone is coming to the App Store
Projects travel in either direction as .zella files. Shoot a property walkthrough on the phone and finish it on the Mac, or rough a long cut on the desktop and review it on the phone. Your timeline, cuts, captions, titles, callouts and effects come across, so nothing is re-done and neither device is a dead end. How the handoff works →
WHEN ONE TAP IS NOT ENOUGH
There is a real editor under the button.
Auto-Polish is a starting point you can accept or take apart. Nothing it does is baked in: every pass lands on the timeline as ordinary, editable, undoable objects. If you have edited before, this is where you find out whether Zella has a ceiling worth growing into.
Free, on every device
- A real timeline - split, ripple delete, drag-trim, drag-to-reorder, per-clip volume, 50-level undo
- J and L cuts, detach and reattach audio, and an Audio tool holding every level in one place
- Zoom blocks, retiming and slow / fast / freeze / reverse speed
- 25 transitions, 43 colour grades on the Mac (42 on iPhone) with an intensity slider, and .cube LUT import
- AI background removal with 40 built-in scenes, subject staging, backlight and a contact shadow
- Caption presets with active-word highlight, 20 callout types on the Mac (19 on iPhone), 16 blend modes and Text Behind Subject
- 43 sound effects on the Mac (30 on iPhone), 16 voice changers, music with Auto-Duck and five studio voice tones
- The AI Assistant and Auto B-Roll, free on both platforms - intelligence is never the paid tier
- Autosave, crash recovery and versioned project snapshots
The Pro depth toolset
Six things separate Pro from Free, and these are the ones you reach for once the work is paid work. A one-time purchase covers Mac and iPhone together.
- Keyframes with a real bezier curve editor
- Live luma scopes and Color Match across shots
- Feathered effect masks, speed ramps and per-drum Beat Sync
- Transparent-background export - ProRes 4444 with alpha on Mac, HEVC with alpha on iPhone
- The multi-aspect export queue
- Saved templates, so your caption style, grade and sound settings are one tap next time
BEFORE YOU DECIDE
The questions people actually ask.
These are the ones that decide it, so they get straight answers rather than marketing ones.
Is my footage uploaded anywhere?
No, on any tier and on every device. Recording, editing, captions, background removal, voice polish and export all run on your own hardware, and the app works in airplane mode. Two things are optional and network-touching if you choose them: the AI Assistant sends your typed prompt to a model using your own API key, never your video, and stock-footage B-Roll fetches clips if you add a free key. The captions model downloads once, about 145 MB, over Wi-Fi.
Do I need an account?
No. There is no sign-up, no login and no profile. Your recordings and projects are ordinary local files that you can move, back up or delete like any others. Nothing expires and nothing is held hostage if you never pay.
Do I have to learn a professional editor?
No, but one is underneath if you ever want it. Record, tap Auto-Polish, export is a complete route to a finished video and you can stop there permanently. The keyframes, curve editor, scopes and masks sit behind an Advanced Tools switch and stay out of the way until you turn it on.
What if I have no editing experience at all?
Then the one-tap path is the whole product for you. It does the work an editor would charge for: cutting the dead air, choosing where to zoom, timing captions to speech, picking and ducking music. If you dislike a result, undo goes back 50 steps, and every automatic pass can be removed on its own without unpicking the rest.
How long does the first video actually take?
For a 60-second talking-head short, a realistic clock is about 5 minutes recording, under a minute for Auto-Polish, 3 to 4 minutes watching it back and nudging things, and roughly a minute to export. Call it 12 minutes from opening the app to having a file. The second one is faster, because the watching-it-back part shrinks once you trust the output.
What does it cost to find out?
Nothing, and there is no card involved at any point. Zella Free is permanent rather than a trial: the complete editor and all the AI, unlimited recording, watermark-free export up to two minutes at 1080p, and 1080p is free forever. On top of that, every new install runs Pro Week - the complete Pro feature set free for seven days, with no card, no account and nothing to cancel. On day eight it simply becomes the free tier again.
See 32 jobs people do with it, and where each one outgrows Free →
Twelve minutes, on your next recording.
Free is permanent, not a trial, and every new install runs the complete Pro feature set for seven days with no card and nothing to cancel. The worst case is that you spend twelve minutes and keep the video.
macOS 14 Sonoma+ · iPhone on iOS 18+ · Local-only · No cloud