How AI clipping picks your moments

Cliperly doesn't guess. Each suggestion is built from measurable signals in your video.

What we measure

• Speech: every word, with its exact timing.
• Energy: how loud each moment is compared with your normal speaking level in this video.
• Laughter and reactions: laughing, exclamations and big reactions in the transcript.
• Your commands: every time you said “clip that”, “clip this” or “clip it”.
• Pace: how continuously you're talking.

How clips are built

Cliperly looks at every run of complete sentences between 15 and 60 seconds long, scores each one, and keeps the best non-overlapping moments. Clips between 20 and 45 seconds get a small bonus, because that length tends to suit short-form feeds.

The Clip Score

The score runs from 35 to 95 and compares moments within the same video. A 95 is the strongest moment in that video — it doesn't mean the clip will get a certain number of views, and Cliperly never claims it will.

Reasons

Every suggestion lists why it was chosen, such as “You said clip that here” or “Voice gets loud and energetic”, so you can judge it quickly.

Local AI titles

Cliperly runs an open-source language model on its own machines to review each candidate, drop incomplete thoughts, and write a hook title, hashtags and keywords. Your footage and transcripts are never sent to an outside AI company.

Five clips per video

Each video gets up to five suggestions. Short videos may produce fewer, because clips never overlap and each must be a complete moment.