Interviews and press conferences become vertical news clips with word-level captions
Whisper writes the transcript on the newsroom's own computer, so unpublished footage stays on your machines.
What slows news clips down
Quotes must be exact
A subtitle that is one word off can change what a source said. Checking it by hand eats an editor's time.
Sources come from everywhere
An official statement on YouTube, a video on X, camera files from a reporter in the field. Formats differ and have to be pulled together first.
The speaker is tiny in frame
Press conference footage is usually wide. Cropped to 9:16, the speaker can end up out of the frame.
A newsroom workflow for vertical clips
- 1
Gather sources
Paste public links from YouTube, X or Facebook, or import camera files and files from the newsroom's Google Drive.
- 2
Transcribe with a larger model
Download the Whisper Medium model in the app for a more careful transcript. Language is detected automatically, foreign-language interviews included.
- 3
Pick the key quotes
Cut sentences straight from the transcript, or have Gemini propose reels in the Educational style with titles in one of 11 output languages.
- 4
Export with separate subtitles
Auto face keeps the speaker centered. Every export also writes an .srt and a caption.txt for the archive and the web version.
Settings for news clips
- Transcript
- Whisper Medium, automatic language detection
- Reframe
- Auto face for a single speaker, fit + blur for wide footage
- Captions
- Active word mode, a plain style with no emoji
- AI style
- Educational
- Export
- Shorts or Reels preset, 1080p, plus .srt
Tools in this workflow
Other use cases
Frequently asked questions

Try it on one of your own long videos
7-day trial on every plan. No credit card.