How to Turn a Long Video Into Short Clips With CapCut AI

The Long video to shorts tool scans a long recording, finds the most engaging moments, and outputs several short vertical clips with captions and reframing already applied. It is the fastest way to mine a podcast, webinar, or stream for social content. This feature lives in CapCut desktop.

What you need

• CapCut desktop, signed in
• A long horizontal video with clear speech (5 minutes or more works best)
• Time for the AI pass, which scales with video length

Step 1: Open the tool

From the CapCut home screen, look in the toolbox for Long video to shorts. Open it and upload your long video, or paste a link if your source is hosted online.

Step 2: Set the clip length and language

Choose a target duration that matches the platform you post to. Set the language to match the speech so captions transcribe correctly. Then click Generate.

Step 3: Wait for the AI pass

CapCut transcribes the whole video, scores segments for hooks and complete thoughts, then assembles candidate clips. A 30-minute source can take several minutes. Each result shows a virality estimate and the topic it covers.

Step 4: Open a clip in the editor

The AI guesses the boundaries, but it is rarely perfect. Click Edit on a promising clip to send it to the full timeline, where you can trim the start so it opens on the hook, and tighten the ending so it does not trail off.

Step 5: Polish and export each clip

Check that the auto reframe keeps the speaker centered, fix any caption errors, then export at 1080p vertical. Repeat for each clip you want to keep.

Result: one 30-minute interview becomes five tightened vertical clips, each captioned and reframed, in roughly the time it takes to watch the source once.