Text-based editing · iPhone
Edit a video by deleting words from its transcript
Finding the edges of a fumbled sentence on a timeline is fiddly on a phone screen. Reading it is not. Transcript editing turns cutting into proofreading: you see what you said, you remove what you do not want, and the picture follows.
The same transcript drives your captions, so fixing a misheard word and cutting a tangent happen in one place.
- iPhone
- Free on every tier
Quick answer
In Zella for iPhone you can cut a video by editing its transcript: generate captions, then delete a word or a line and Zella ripple-cuts the video to match. The transcript comes from a verbatim on-device Whisper model, so even um and uh are there to delete. It is free on iOS 18 or later, and nothing is uploaded.
How to do it in Zella
-
Step 1: Add captions
Open the clip in the editor and add captions from the Captions tool. Zella transcribes on the phone into a word-level transcript.
-
Step 2: Open the transcript
Open the transcript word editor, where every word is listed in order with its exact timing.
-
Step 3: Delete what you do not want
Delete a word or a whole line. Zella ripple-cuts the video at those words and closes the gap.
-
Step 4: Correct misheard words
Tap a word to fix its spelling. The timing stays intact, and the fix shows in the captions and in the export.
-
Step 5: Tighten the rest in one tap
Tap Clean Up to remove the remaining silences and filler words, then export.
Why do it this way
Cut by reading
The quickest way to drop a fumbled sentence or a tangent is to find it in the text and delete it, not to hunt for it in a waveform.
A verbatim transcript
iOS transcribes with an on-device Whisper model that writes disfluencies out in full, so every um and uh is visible and removable.
One source for cuts and captions
Captions come from the same transcript, so they stay on the right words after every cut you make.
What you need
- iPhone on iOS 18 or later
- The transcription model downloads once, about 145 MB, the first time you transcribe; after that it works offline
- Speech in a language Zella transcribes: 27 are offered, 18 verified end to end on iPhone
- Free on every tier: no watermark, exports up to 2 minutes at 1080p
Questions
Is transcript editing in the Mac app?
Deleting words to cut the video is an iPhone feature. On the Mac you cut on the timeline, and Cut Empty Space and Remove Fillers handle silences and fillers in one click each. An iPhone project opens on the Mac as a .zella file.
Is my video uploaded for transcription?
No. Transcription runs on the phone. The only network step is the one-time model download of about 145 MB.
Which languages can I edit in?
Zella transcribes 27 spoken languages, and 18 are verified end to end on iPhone, including English, Spanish, French, Hindi, Arabic and Japanese. Bangla and Telugu are not transcribed on-device.
Can I undo a cut I made in the transcript?
Yes. Transcript cuts are normal edits, so the editor's Undo control brings the words and the video back.
Related tools
Read the manual