How a long video becomes vertical clips

Clippi always follows the same path, and every step is visible in the Studio: status, logs, validation criteria. You decide on the selection, the render and each post.

  1. 1. You import a video you hold the rights to

    A file (MP4, MOV, MKV, WebM, AVI) or a direct address. You confirm you hold the rights: the confirmation is kept with the project. Clippi does not download from YouTube, whose terms forbid it; you import the file of your own video.

  2. 2. Clippi transcribes word by word

    The timestamped transcript is the basis for everything else: finding passages, synchronised subtitles, cuts at the end of a sentence.

  3. 3. It suggests the passages that stand on their own

    Each passage gets an editorial score explained criterion by criterion (hook, self-containment, interest, pace, length). The score measures the quality of a passage; it does not predict views.

  4. 4. You choose, and you see the cost first

    Rendering time and cost are shown before anything starts. Nothing is produced without your approval of the selection (unless you choose autonomous mode).

  5. 5. It produces each clip

    Cut, vertical framing on the face, audio processing, subtitles, keywords, title and description. A single final encode at 1080×1920.

  6. 6. Every clip goes through a quality check

    Format, audio/video sync, volume, black frames, subtitle legibility, mid-sentence cuts, redundancy, title consistency, media rights. A rejected clip goes back to the responsible agent, with at most two automatic corrections.

  7. 7. You fix and schedule

    An editor with in and out points, text, style, music. Each change creates a new version; the old one can still be restored. Posts go through a schedule with per-account quotas, and each one waits for your approval.

The ten agents

Clippi Core is in charge. Each agent has its inputs, outputs, permitted tools and validation criteria.