UGC testimonial ad with an avatar: 10 variants of 15 s cost $27.60
A 15-second talking-head UGC ad costs $2.76 on Sume's Standard avatar tier, $3.68 on Plus and $8.25 on Max. Ten script variants cost $27.60, $36.75 or $82.50.
A 15-second talking-head UGC ad costs $2.76 on Sume's Standard avatar tier, $3.68 on Plus and $8.25 on Max. Ten script variants cost $27.60, $36.75 or $82.50 respectively, before voice and captions.
Use this route when someone must speak to the camera. A video generation model such as Gemini Omni Flash 1.1 takes image and video references only, with no audio input, so it cannot speak a script you wrote.
Cost by tier
Avatar video is billed per second of output, and a script must come out between 4 and 60 seconds. The job needs a ready avatar, which is created separately from the clip.
| Tier | Per second | One 15 s ad | 10 variants |
|---|---|---|---|
| Standard | $0.184 | $2.76 | $27.60 |
| Plus | $0.245 | $3.68 | $36.75 |
| Max | $0.550 | $8.25 | $82.50 |
What to vary across ten variants
The face is the same in every variant, so the script does the work. Keep each variant to one change.
- Opening line: a problem statement, a number, or a question.
- Proof: a result, a before-and-after, or a plain description of use.
- Call to action: the last three seconds.
- Length: 15 seconds against 10, since a shorter script is also cheaper per variant.
Cost levers
Length is the lever you control. Ten variants at 10 seconds on Standard cost $18.40, against $27.60 at 15 seconds. A tier change is the other: Plus is 1.33x Standard per second and Max is 2.99x.
Pick the tier on a single test clip. Pay for a higher tier only where the difference shows on a phone screen at the size viewers will see.
Be clear with viewers
An avatar is not a customer. Do not write a script that claims the speaker bought and used the product. Write it as a presenter explaining the product, and follow your ad network's rule on synthetic people.
Captions are a separate $0.20 job on video captions. Rates are from the Sume API price card on 2026-10-07, and the avatar docs hold the live request shape.
Sources
Related posts
More in Sume Avatar 1.0
- What a scripted AI avatar clip cannot do: seven limits and fixes
A Sume avatar clip cannot take questions, run past 60 seconds or switch presenter mid-video. Seven documented limits, each with a workaround.
- Introducing Sume Avatar 1.0
Sume Avatar 1.0 is a multi-agent orchestration system as a single avatar model.
- Avatar Face Swap API (Beta): apply an avatar face to a video
Avatar Face Swap 1.0 is a Beta Sume endpoint that applies a ready avatar's face to a short public source video. Required fields, limits, and polling.
- Avatar video previews: approve the first frame before rendering
Create an avatar video preview to get first-frame stills, regenerate them if needed, then call generate-video on the preview id to render the final video.
Written by Sume