Anime-style ad with Seedance 2.5: use a character sheet as references
Send your mascot's character sheet as input_references to seedance-2.5 on Sume and test a 6-second anime-style ad at 480p before you pay for 720p.

To keep an illustrated mascot on-model across a short ad, send the character sheet as images in input_references and describe the action in the prompt. On seedance-2.5 a 6-second test costs about $1.61 at 480p on Sume, and about $3.47 at 720p.
References are guidance, not an exact copy. Expect the outfit and face to be close, and plan to reject takes where they drift.
Build the reference set
Use two to four images: a front view, a three-quarter view, a color palette card and, if you have it, a background style frame. Each is one image_url entry. Seedance 2.5 takes many references on the provider side, but check the limit on the Sume catalog for the id before you send a large set (video generation docs).
{
"model": "seedance-2.5",
"prompt": "The mascot runs through a rainy neon street, hand-drawn anime style, flat colors",
"duration": 6,
"resolution": "480p",
"aspect_ratio": "9:16",
"input_references": [
{"type": "image_url", "image_url": {"url": "https://example.com/mascot-front.png"}},
{"type": "image_url", "image_url": {"url": "https://example.com/mascot-three-quarter.png"}},
{"type": "image_url", "image_url": {"url": "https://example.com/palette.png"}}
]
}Price
Seedance 2.5 is billed per video token times 1.25 on Sume. The rate card charges output tokens (width x height x seconds x 24 / 1024) at $0.0214 per 1,000 tokens at 480p and 720p, times 1.25, so 6 seconds is about $1.61 at 480p and about $3.47 at 720p.
| Resolution | Approx. price | Five takes |
|---|---|---|
| 480p | about $1.61 | about $8.05 |
| 720p | about $3.47 | about $17.35 |
What to judge in a take
Check three things on each result: whether the hair and outfit colors match the palette card, whether the face stays on-model when the character turns, and whether the line weight stays consistent. ByteDance says Seedance 2.5 gives smoother, more consistent motion (read 2026-10-07, Seed), but you should confirm that on your own character rather than rely on the claim.
Rights
Only send characters you own or have a license to use. Do not ask for an existing studio's characters or style by name.
Practical notes
Keep the action simple in the first test: one character, one location, one movement. Once the character holds up, add a second reference for a prop or a background style and run a longer take.
Write the style once and reuse it. A line such as hand-drawn anime style, flat colors, clean outlines, repeated word for word across jobs, does more for consistency than rewording the prompt each time.
Sources
Related posts
More in Models
- Can Gemini Omni Flash speak my script? It takes no audio input on Sume
On Sume, Gemini Omni Flash 1.1 takes image and video references but no audio reference. It makes its own sound. For your script in a face, use avatar video.
- Same character in three AI video shots: Omni reference image
Keep one character across separate Gemini Omni clips by sending the same reference image with every request on Sume. Request shape, prompt tags, 3-shot cost.
- ChatGPT Image 2 vs 2.5 at medium quality, 16:9: price on Sume
At 16:9 and medium quality, ChatGPT Image 2 bills $0.0525 per image on Sume and ChatGPT Image 2.5 bills $0.0105: five times less. Table, batch math and a swap.
- Do low, medium and high mean the same on every Sume image model?
No. On Sume only ChatGPT Image 2, 2.5 and Ideogram 4.5 accept low, medium and high, and the tiers change price in different ways. Defaults and the tier ladders.
Written by Sume