Skip to content
CutSpool

Use case

Find the compact answer inside a long interview.

An interview runs long and the answer worth posting is short. The recording is transcribed with word-level timings, the ranges come back scored, and each one is cut to vertical 9:16.

Start clipping
Source

Upload the recording as it was captured.

MP4, MOV or WebM, within your plan's size and length ceilings. Nothing is processed while the file is uploading; the job is queued once the upload is verified in storage.

The upload panel, with a source file staged and the run settings on the right.
Transcript

Question and answer, timed to the word.

The audio is transcribed with word-level and segment-level timings, and every cut is snapped to a word edge so a clip does not open halfway through a sentence. The source transcript stays inside the pipeline; while the source is retained, the clip transcript can be corrected in the editor before you post it.

The same workspace while a run is still cutting.
Candidates

The short version of a long answer.

Candidates arrive as ranges with a title, a description, hashtags and a score from 0 to 10, ordered highest first. When one answer is spread across the conversation, a candidate can join up to four ranges into a single cut.

One finished clip as it appears in the list.
Output

One answer, vertical, captioned.

Clips render vertical at 9:16, 1080 x 1920 or 720 x 1280. Captions are burned in when enabled and can be hidden before re-rendering and downloading. While the source is retained with READY status, the editor can change the caption text and style, framing, burned-in title, or shorten and re-extend the clip inside its original generated timeline. The model-selected source moment cannot be replaced, and source-backed editing ends when the source expires or is deleted.

The clips workspace, rendered by the app's own components with a sample run.

Sample run

What a 48-minute conversation returned.

One interview file, one run, three scored candidates, two of them stitched from more than one moment in the conversation. Every figure below is read off the run the clip list is drawing.

Source file
interview-nadia-rademaker.mov
Length
47:54
Credits charged
48
Clips returned
3
Ranges cut
4 across 3 clips
Highest score
8.9
That clip's title
The first thing she changed was who answers the phone
Start clipping
The clips workspace, rendered by the app's own components with a sample run.

The parts that matter here.

AI clipping
Read
Transcription
Read
Reframe
Read

What a run costs.

Credits are counted off the interview you upload, never off the answers it yields. The 47:54 recording in the sample run above came to 48 credits, and the three clips it returned added nothing on top of that.

See pricing

Other recordings this works on.

Podcasts

Turn a 60-minute episode into a queue of reviewable vertical cuts.

View podcast workflow

Webinars

Pull reusable teaching moments out of a long session.

View webinar workflow

Creator content

Get a week of short posts out of one long upload.

View creator workflow

Business content

Cut the recordings your team already has into short clips.

View team workflow

Interview questions.

The original range comes from the model, so sometimes a clip is the answer alone and sometimes it opens with the question that prompted it. There is no arbitrary source-range picker. While the source is retained, you can shorten the generated clip or re-extend it inside its original timeline.

Yes. A candidate is one to four ranges from the same recording, stitched in playback order, so an answer returned to twice in an hour can arrive as one cut. The model decides that, and only when the ranges share one theme.

Your plan sets the ceiling for a single file, from 60 minutes at the lowest to four hours at the highest, and a file past it is refused before anything is charged. Length on its own is not a wait problem: the audio is transcribed in five-minute chunks running in parallel.

On the plans that carry it, yes, and it is set before the run. Pause removal drops silences longer than 1.5, 2 or 3 seconds, whichever you pick. Tightening removes long pauses and hesitation sounds, then plays the clip 8% above conversation speed. Whatever you choose applies to every clip in that run.

Yes, while the source is retained with READY status. The editor shows the clip's own transcript on its timeline, and any word can be corrected - useful when a name or a technical term came back wrong. A correction changes the text that is burned in; it does not choose a new source moment.

Ready when the recording is

The answer worth posting is already in the recording.

Start free