AI video defaults on Sume: 768p, 480p, 720p and 8 seconds explained
Recast defaults to 768p, Genjutsu to 480p, Gemini Omni Flash and sume/auto to 720p and 8 s. Which defaults change price, and the one place fal differs.

When you leave resolution or duration off a Sume video request, the row fills in a default. Most are sensible, but a few differ from what you might assume from the vendor's own page, and a default that picks a higher tier costs more. The catalog documents four that matter.
Which defaults does Sume set?
The values come from the Sume Video Router constraints, not from this post, so read the live catalog if you are unsure.
| Row | Default | Note |
|---|---|---|
h3-max-recast | 768p | fal's own default is 1080p; Sume sets 768p, which is cheaper at $0.375 per second |
higgsfield-genjutsu | 480p | Duration is the inspected source length, rounded up |
gemini-omni-flash-1.1 | 720p, 8 s | Edit mode: resolution optional, default 720p, no duration sent |
sume/auto | 720p, 8 s | Validated against the Gemini Omni Flash 1.1 envelope |
Why does the Recast default matter?
Recast at 768p is $0.375 per second billable; at 1080p it is $0.5625, a 50 percent step. A script ported from fal's docs, which default to 1080p, would expect the higher tier. On Sume you get 768p unless you ask for 1080p, so a first run costs less and looks softer. Send resolution: "1080p" explicitly when you need it.
What should I do about defaults?
Do not rely on them in code that has a budget. Send resolution and duration on every request so the cost is visible in your own request log, and so a catalog change cannot move your bill. For Genjutsu, defaulting to 480p keeps trial runs cheap at $0.3975 per second, but the 720p tier is $0.85125 per second, so a quality check there is more than double.
The same logic applies to Auto: leaving out duration gives eight seconds, which at 720p is a fixed $1.00 per clip at Gemini Omni Flash 1.1's rate.
Sources
Related posts
More in Developers
- AI video models with no aspect_ratio option: Recast, Genjutsu, Grok
Three Sume video rows reject aspect_ratio: H3 Max Recast, Higgsfield Genjutsu and Grok Imagine Video 1.5. What sets the output frame instead, and what to send.
- AI video person swap rejects long takes: find shots over 15 s
H3 Max Recast on Sume needs 5 to 30 seconds with no single shot over 15. Find the long takes with ffmpeg scene detection, then cut them before you pay.
- AI voice reads Spanish with an English accent: two causes to check
Spanish read with an English accent usually means the language field was omitted or the voice is tagged for another language. How Sume's TTS 409 works.
- Amazon Sponsored Brands video: Sume fps 24-30 fit; 60 and 4K do not
Amazon lists 23.976 to 30 fps and 1280x720, 1920x1080 or 3840x2160. Sume trim offers 24, 25, 30 or 60 fps and stops at 2160 per edge. Which settings line up.
Written by Sume