Kids story, five 8-second scenes: Omni $5.00 vs H3 $3.00

A story of five 8-second scenes is 40 seconds of footage, and it costs $5.00 on Gemini Omni Flash 1.1 at 720p, $3.00 on MiniMax H3 at 768p and $4.00 on H3 Max at 768p on Sume. Per scene that is $1.00, $0.60 and $0.80.

The cheaper model is not automatically the right one for a story, because the same character has to show up in every scene. Prices come from the Sume price tool (read 2026-10-09).

The five-scene price

Omni takes up to 10 reference images per request. You can hold the character sheet in the reference list for every scene and change only the prompt.

One character, five requests

Keep the same reference image list in all five bodies. Put the character first so the token meaning never moves, and write the scene after it.

{
  "model": "gemini-omni-flash-1.1",
  "prompt": "Scene 3. <IMAGE_REF_0> carries a lantern through <IMAGE_REF_1>, night, gentle camera follow",
  "reference_image_urls": [
    "https://example.com/fox-character.png",
    "https://example.com/forest-background.png"
  ],
  "resolution": "720p",
  "duration": 8,
  "aspect_ratio": "16:9",
  "mode": "async"
}

Longer and shorter stories

The price of a story is scenes times the scene price. Three scenes is a trailer; eight is a short.

Limits and gotchas

Dub narration separately: both models always produce native audio, and neither has a silent rate.

• Omni scenes are 3-10 seconds; H3 scenes are 5-15 seconds.
• H3 accepts up to 9 reference images, 3 videos and 3 audio files, 12 in all. The first five reference images on H3 are free and each extra one adds a charge, so a character sheet of six or more costs more on H3.
• Give each scene its own Idempotency-Key, such as fox-story-scene-3-v1.
• The scene ends when the clip ends. If a line of narration is longer than 8 seconds, split the scene.