Step 1 · Ingest

Upload a base video and its AI-generated variants.

Every file is decoded with OpenCV at one frame per second, then frames are aligned by timestamp so 00:03 in the base maps to 00:03 in each variant.

Base Video

Drop a video or click to browse

The original master before any swaps

.mp4 · .webm · ≤ 100 MB

Variant 1

Drop a video or click to browse

Required · a swapped person, product or scene

.mp4 · .webm · ≤ 100 MB

Variant 2

Drop a video or click to browse

Required · a swapped person, product or scene

.mp4 · .webm · ≤ 100 MB

Variant 3

Drop a video or click to browse

Optional · add up to four variants

.mp4 · .webm · ≤ 100 MB

Variant 4

Drop a video or click to browse

Optional · add up to four variants

.mp4 · .webm · ≤ 100 MB

0 base ·0 / 4 variants(min 2) .mp4 / .webm · ≤ 100 MB · ≤ 60 s each