18-second UGC-style ad with a product image: $3.49, $4.64 or $10.44

An 18-second UGC-style ad made with Sume Avatar 1.0 and a product image costs $3.49 on the standard tier, $4.64 on plus and $10.44 on max. Without a product image the same clip is $3.31, $4.41 or $9.90. Plus is the default when you omit quality, so say which tier you want.

The rates come from Sume's Avatar Video price table: $0.184, $0.245 and $0.55 per second without a product image, and $0.194, $0.258 and $0.58 with one. The Avatar Video docs (read 2026-10-06) describe the three tiers and the 4 to 60 second window; the arithmetic below is straight multiplication, so check the estimate on a preview before you commit to a large batch.

The 18-second price table

UGC-style ads are short, vertical and cheap to test, and 18 seconds is long enough for a hook, one product beat and a call to action. At Sume's estimate of about 2.8 spoken words a second, that is a script of roughly 50 words.

Which tier for which job

Standard is the fastest path in Sume's own description, so it suits hook tests where you will throw most of the clips away. Max is the slowest and the most expensive, so it earns its place only on the one clip you will put money behind.

A practical pattern is to write three scripts, render all three on standard, and re-render only the winner on plus or max. The preview flow helps here: you create an Avatar video preview, approve the first-frame stills, then call generate-video on the preview id. The docs say a quality override at generate-video changes only the final render tier, and the preview stills are reused, so approving at one tier and rendering at another does not need a new preview.

What the product image changes

The product image costs a little more per second: $0.01 on standard, $0.013 on plus and $0.03 on max. Over 18 seconds that is 18 cents, 23 cents and 54 cents. The image must be a fetchable public HTTPS URL; Sume rejects other schemes with a 400.

Two limits to plan around. The current execution supports one resolved avatar per final video, and the resolution is 720p. If your ad needs 1080p or two on-screen speakers, this is not the surface; say so before you commit to a campaign built around it. Captions are separate: inline captions burn into the clean final MP4 and are a separate add-on in the avatar estimate, so they are not in the table above.