PricingWhen to use it
For talking-head videos, interviews, founder updates, tutorials, podcast clips and UGC ads, anywhere a speaker’s hesitations slow the video down. It suits anyone who would rather describe the edit than hunt for each “um” on a timeline.
Inputs, decisions and output
| Part | Details |
|---|---|
| You provide | Footage with speech, and one line in your brief or a revision, such as “remove the ums and false starts.” |
| Shorty handles | A verbatim transcript that keeps every hesitation, phonetic alignment of each word to the audio, a judgement on each candidate in its sentence, and the cuts in picture and sound. |
| You control | What to cut and what to keep: only ums and uhs, or tics such as “like” and “you know” as well, and where the take should start. |
| You receive | The finished edit built on the clean take, with captions, sound cues and length timed to the speech as it is now. |
How to use it
- Add the clips with speech to Shorty.
- Write the rest of your brief as usual, and add a line such as “remove the ums, uhs and false starts.”
- Say what to keep if it matters, for example “leave ‘like’ in, that’s how I talk.”
- Submit, then watch the result. Ask in a revision if a cut needs to change.
Example instruction
Adapt this instruction to your footage and review the result.
Make a 45-second vertical cut of this interview. Remove the ums, uhs, stutters and false starts, keep my pauses natural, and add word-by-word captions.
See it on four real talks
We ran four public talks through the same tool the director uses inside an edit. Across them, 107 words came out in 80 cuts, including every “um” and “uh” in the transcripts, and every kept word survived in order. The launch post plays each one before and after, with every removed word struck through.
| Recording | Length before → after | Words cut | Cuts | Ums and uhs left |
|---|---|---|---|---|
| Nik Storonsky, Revolut | 2:00 → 1:49.8 | 40 | 28 | 0 of 13 |
| Prime Intellect, NVIDIA event | 1:00 → 0:53.2 | 38 | 27 | 0 of 16 |
| Arjun Karanam, Trajectory | 2:00 → 1:54.4 | 22 | 21 | 0 of 8 |
| Elon Musk, Tesla Roadster event | 2:00 → 1:58.4 | 7 | 4 | 0 of 1 |
Sources: Shorty News: four talks before and after. Measured on excerpts of public YouTube talks on 20 October 2026. Results depend on the recording.
What to check
Listen at each cut. Joins are cut at sound boundaries with a 4-millisecond crossfade, but a cut can still be visible as a small jump on camera; ask for B-roll or captions over it in the same brief.
A hesitation the recognizer doesn’t hear can’t be cut. English and Russian are tested by ear today; French, German, Spanish, Portuguese, Italian and Dutch are in testing. Each clip can have up to three minutes of speech, on up to four clips per edit, and Shorty never removes more than 30% of a recording as filler.