Nectr

Sign in

This account was created for one person. If that is not you, it will not let you in.

Nectr
Your account

Welcome, .

Here is the engine on our own footage — 41:07 of raw kitchen video, cut to 1:08 without a human touching the timeline. Below it, tell us what you want it pointed at.

41:07
source
1,234
frames read
14
moments kept
97.2%
discarded

What it produced

Vertical 1080×1920, reframed from 16:9 by tracking where the motion is, captions written from what the model saw, mastered to −14 LUFS so it plays at the same volume as everything else in a feed.

01Every 2 seconds of footage is read by a vision model on our own hardware.
02Moments are scored on focus, clarity and how much of the surrounding footage agrees.
03The cut, the captions and the crop are all derived from those scores — no template.
04If the footage is not something it handles well, it refuses instead of guessing.

Full breakdown of every decision, with the frames: see the analysis.

Your first job

Tell us what to point it at.

We tune each vertical by hand before it runs properly. Tell us what you shoot and which platforms you need, and we will come back to you with what we can do.

Upload

Give it an episode.

A file from your machine, or a link to one already published. It goes straight to our own hardware, and you see each stage as it happens — the same stages we see.

OR PASTE A LINK