26-second AI video API: estimate the price before you submit
A 26 s Wan 3.0 clip bills three amounts on Sume. Read pricing_skus, reserve and refund rules, and check duration support before you submit.

A 26 second clip on Sume can come from Wan 3.0 or Seedance 2.5. For Wan, the bill is $1.63 at 480p, $3.25 at 720p or $6.50 at 1080p. The way to avoid surprises is to read pricing_skus from the catalog before you submit.
Which models accept 26 seconds
Seedance 2.5 is billed per video token rather than per second, so a per-second figure would be misleading; the catalog row gives the rate.
| Model id | Duration range | Billed for 26 s |
|---|---|---|
| wan-3.0 | 2-30 s | 480p $1.63, 720p $3.25, 1080p $6.50 |
| seedance-2.5 | 4-30 s | token priced, read pricing_skus |
| higgsfield-genjutsu | 4-30 s (source length) | read pricing_skus |
| h3-max-recast | 5-30 s (source length) | 768p $9.75, 1080p $14.63 |
Reserve, settle, refund
When you submit, Sume reserves the full amount for the requested duration and resolution. If the job fails or is cancelled, the reservation is released. This means a 26 s request needs enough balance for the whole clip up front.
Omitting duration reserves each model's default, which is why the stored post on omitted resolution and duration matters for budgeting.
What a 26 second request costs
Sume bills the provider list price times 1.25, rounded up to the next cent. The amount is reserved when you submit and refunded if the job fails, so a rejected or failed clip does not cost you. The /v1/videos/models descriptor lists the billable rate in pricing_skus, which is the field to trust over any number in a blog post.
Draft a 26 s clip at 480p for $1.63, then pay $6.50 once for the 1080p final.
Check 26 seconds against the live catalog
Duration support lives in the catalog, not in this page. Each entry on GET /v1/videos/models carries supported_durations, which is every whole second from the model minimum to its maximum, so one membership test answers the question. This script lists every model that accepts 26 s, with its resolutions and billable rates.
Set SUME_API_KEY first. See the Video Generation docs for the request and polling lifecycle.
import os
import requests
N = 26
resp = requests.get(
"https://api.sume.com/v1/videos/models",
headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"},
timeout=30,
)
resp.raise_for_status()
for m in resp.json()["data"]:
if N in m["supported_durations"]:
print(m["id"], m["supported_resolutions"], m["pricing_skus"])Sources
Related posts
More in Pricing
- A 30-second Omni ad cost sheet: three clips, a Lyria bed, a Timeline
Three 10-second Omni clips, one Lyria 3.5 track and a Timeline render: about $3.08 on Google's list prices and $3.98 on Sume, before your own assembly work.
- 3D model API prices: Meshy, Tripo, Rodin side by side
3D model APIs run from about $0.01 per credit to $0.50 per model. Sume has no 3D model endpoint; compare per-asset cost with a still image or short spin clip.
- 4-second AI video API: which Sume models take 4 s and the price
Eight Sume video ids accept a 4-second clip; MiniMax H3 does not. Price per 4-second job from $0.05 to $2.32 by model, with the 720p row for each.
- 40-second AI video API: a 30 s job plus a 10 s job on Sume
A 40-second AI video on Sume is a 30-second job plus a 10-second job. Wan 3.0 costs $2.51, $5.00 or $10.00 at 480p, 720p and 1080p. Split, price, join.
Written by Sume