10 two-second ad hooks on Wan 3.0: $2.50 at 720p on Sume

Ten two-second ad hooks cost $2.50 at 720p on Wan 3.0, the only model in Sume's video catalog that accepts clips shorter than 3 seconds (read 2026-10-08). At 480p they cost $1.30 and at 1080p $5.00. Every other model starts at 3 seconds (Gemini Omni Flash 1.1) or 4 to 5 seconds.

Minimum length by model

A hook is the first beat of an ad, so a 2-second test is closer to the final cut than a 5-second clip you trim down. Sume bills a clip as its own job, and billable prices are the catalog list times 1.25 rounded up to a cent.

Ten hooks, three resolutions

Wan 3.0 accepts 2 to 30 seconds at 480p, 720p and 1080p. It takes up to 10 reference images, so you can hold the product constant across hooks. It accepts 9:16 among its aspect ratios, and its price does not change with the ratio.

Attach each hook to one body

Join a hook to a shared body on Timeline 1.0. Keep one voiceover as the spine and use audio.source_in and video[].source_in to cut. A 25-second ad is still one billed minute, so each finished variant adds $0.10 for the render. Ten variants therefore add $1.00 on top of the hooks, and the whole test at 720p is $2.50 + $1.00 = $3.50.

curl -X POST https://api.sume.com/v1/video-router/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: wan-hook-04" \
  -d '{
    "model": "wan-3.0",
    "prompt": "Close-up of the product dropping onto a table, sharp sound-ready impact frame",
    "resolution": "720p",
    "duration": 2,
    "mode": "async"
  }'

When 2 seconds is too short

Two seconds leaves little room for speech, so treat it as a visual hook and let the voiceover start on the join. If your hook needs a spoken line, move to the 3-second Omni Flash row, where audio is always generated.