# How do you turn a long video into Shorts?

> Upload a podcast, stream or talk. The AI clip maker finds 3-10 moments that open on a hook, frames the speaker in 9:16 and captions every word. You review and trim each clip before it is made.

Canonical: https://faceless.so/tools/ai-clip-maker

AI clip maker

To turn a long video into Shorts, upload it (MP4, MOV, WebM or MKV, up to 2 hours). Faceless transcribes it with speaker labels, picks 3 to 10 moments of 20-60 seconds that open on a hook and end on a finished thought, reframes each shot to 9:16 around the people in it and captions every word. You review, trim and fix the captions, then each clip you keep becomes its own editable video. Finding clips costs 15 credits per started 10 minutes of video; each clip you make costs 8 credits per started 30 seconds.

Upload up to 2 hoursClips start on a hookSpeaker kept in frameCaptions word by wordReview before you pay

Updated November 21, 2026

## Upload a podcast, stream or talk

MP4, MOV, WebM or MKV, 2 minutes to 2 hours and up to 6 GB, that you own or have the rights to clip.

[Upload your video](https://faceless.so/create?type=clips)15 credits per 10 min to find clips, 8 per 30 s clip made

## How to turn a long video into Shorts in 5 steps

1.  1
    
    ### Upload the video
    
    Upload an MP4, MOV, WebM or MKV up to 2 hours that you own, and confirm you have the rights to clip it.
    
2.  2
    
    ### Choose the clips
    
    Pick how many clips (3-10), short or long clips, the caption style and, if you like, what they should be about.
    
3.  3
    
    ### Let it find the moments
    
    It transcribes the video, picks the moments that open on a hook and reframes each one to 9:16. A one-hour podcast takes a few minutes.
    
4.  4
    
    ### Review and trim
    
    Play each clip, trim the start and end, fix a caption word, keep the ones you like.
    
5.  5
    
    ### Make and post
    
    Each clip you keep becomes its own video in the editor and renders to an MP4 you can download, edit further or schedule.
    

## What makes a clip worth posting

-   ### It opens on a hook
    
    The first sentence works alone: a claim, a number, a question or the start of a story. Never "yeah, so" or a reply to something the viewer did not hear.
    
-   ### It ends on a finished thought
    
    The point, the punchline or the answer is said before it cuts. Clips are cut on sentence boundaries, never mid-word.
    
-   ### The speaker stays in frame
    
    The vertical window follows the camera cuts and the person in each shot. A wide two-shot becomes a split screen with both people. Slides and screen shares are shown whole over a blur.
    
-   ### Every word is captioned
    
    Captions come from the same transcript, timed word by word, in a style you pick, without the "um"s, "uh"s and stutters (the sound keeps them). Fix any word and it keeps its timing.
    

## How it compares with dedicated clipping tools

Dedicated clippers such as OpusClip, Vizard or quso are built around clipping alone and offer more there: virality scores, brand templates, team review and many export presets, and most of them also import from a YouTube link. If clipping a weekly podcast is the whole job, one of them may suit you better.

Faceless is a studio for running channels, and the clip maker is one way in. What is different: you see and trim every clip and its price before anything is charged for making it; each clip becomes a full editor project where you can add b-roll, music or AI scenes; and the same account makes faceless videos, series and long-form, and posts them on a schedule. It only clips videos you upload and own.

## Frequently asked questions

Everything about turning long videos into clips

### Can I paste a YouTube link instead of uploading?

No. The clip maker works from a video file you upload and own (or have the rights to clip), and it asks you to confirm that. It does not download videos from YouTube or any other site. If the video is yours, download your own copy from YouTube Studio and upload that file.

### How much does it cost?

Making clips needs a plan, from $29/mo. Finding the clips costs 15 credits per started 10 minutes of video (a 1-hour podcast is 90 credits), charged when you start; if that step fails, the credits come back. Each clip you choose to make costs 8 credits per started 30 seconds (8 for a 30-second clip, 16 for a 60-second one). A 1-hour podcast cut into eight clips is about 186 credits. Both prices are shown before you tap, and they are the prices charged.

### How does it pick the clips?

The whole video is transcribed with speaker labels, then an AI editor reads the numbered transcript and names the moments that work alone: the first sentence has to stop a scrolling viewer (a bold claim, a surprising fact, a sharp question, the start of a story), and the last sentence has to finish the thought. Clips are cut on sentence boundaries, so none starts or ends in the middle of a word, and each comes with a title and a one-line reason.

### Does it keep the speaker in frame?

Yes, with one limit. Every clip is split at the camera cuts and the faces in each shot are found. One person in the shot: the 9:16 window is centred on them and re-centres when they move. Two people close together: the window sits between them. A wide shot with two people too far apart for one vertical frame is usually shown as a split screen, one person on top and the other below, so both stay in frame; it frames only the person talking when it can tell from their mouth movement who it is, which a wide shot rarely shows clearly. Three or more people far apart, or no face at all (a screen share, slides): the whole frame is shown across the width over a blurred copy of itself, so nothing is cut off. You can change the framing of any clip in the editor.

### Can I edit the clips?

Before anything is made you can trim each clip's start and end, rename it, fix any caption word (the fixed word keeps the timing of what was said) and pick the caption style: HORMOZI 1, Iman, MR BEAST 1, Highlight, Ali, DEVIN, POP, Noah. After you make them, each clip opens in the Faceless editor like any project, where you can change the framing, the captions, add music or b-roll, and export or schedule it.

### What videos work best?

Talking videos: podcasts, interviews, livestreams, webinars, talks, coaching calls and vlogs, between 2 minutes and 2 hours long. A clear voice matters more than the camera: the clips are found from what is said. Music videos and silent footage have nothing to transcribe.

## More free tools

[YouTube Shorts safe zone See what the Shorts buttons cover before you post.](https://faceless.so/tools/youtube-shorts-safe-zone)

[Long-form AI video generator A narrated 16:9 video from one topic.](https://faceless.so/tools/long-form-ai-video-generator)

[Lyric video maker Your song as a lyric video, synced word by word.](https://faceless.so/tools/lyric-video-maker)

[Audio to video A free waveform video of a track, made in your browser.](https://faceless.so/tools/audio-to-video)

## Ready to Create Your Own Videos?

Start creating amazing AI-powered faceless videos in minutes with Faceless

[Start my channel](https://faceless.so/projects)[View Pricing](https://faceless.so/pricing)

Free to sign upPlans from $29/moCancel anytime
