Formats

Every deliverable derives from one timeline, so times always match.

What you get

MP4: 1080p, 16:9, chapters embedded, one file per language version.
WebVTT: Sidecar files for captions and chapters.
JSON: The step list with ids, timestamps, titles and screens, for help-center integrations.
YouTube: A chapter list to paste into the video description.

One entry per step:

vid_9f2k.manifest.json

{  "video": "vid_9f2k",  "language": "en",  "steps": [    { "id": "open-invoice", "t": 12.4, "title": "Open the invoice", "screen": "/invoices" }  ]}

Chapter marks are module boundaries. They are generated from the video's structure, never set by hand.

Updated 2026-08-05