Amazon holiday video: 15-30 seconds, legible muted, captions for $0.20
Amazon Ads advises 15-30 second holiday videos that work without sound. On Sume, speech clips caption from audio and silent ones from cues, $0.20 a job.
Amazon Ads' holiday guide says to keep the video brief and focused at 15 to 30 seconds, and says videos should work without sound and stay legible on mobile (Amazon Ads guide, read 2026-10-05). The practical fix is burned-in captions: on Sume, POST /v1/video-captions is $0.20 per job whether it reads speech or burns text you supply.
Two ways to caption, one price
A clip with audible speech is captioned from its audio: send video_url and, optionally, script_text to align your exact wording. A clip with no speech, such as a music-only product spot, fails with caption_no_speech, and the error's next_action is use_overlay_captions. For that clip, send authored text as cues (or segments), each with text, start, and end.
Both paths are one job at the same flat price, so the choice follows the clip rather than the budget. Note that the input is a public HTTPS video URL, unlike trim and Timeline, which take media.sume.com artifacts.
| Clip | Audio | Caption input | Price |
|---|---|---|---|
| 15 s product demo with voiceover | Speech | Speech-to-captions, optional script_text | $0.20 |
| 20 s music-only gift montage | None | cues with text, start, end | $0.20 |
| 30 s unboxing with talk | Speech | Speech-to-captions | $0.20 |
| 15 s silent product loop | None | cues; no cues means caption_no_speech | $0.20 |
Why captions are not optional here
Amazon's advice to make the video work without sound is a statement about how holiday ads are consumed: scrolling, often with the volume off. A voiceover that carries the offer is invisible in that moment, so the offer has to be on screen as text. Treat captions as the primary carrier of the price and the deadline, and the audio as the second layer.
What Amazon's other advice adds
The same guide says savings claims must be identical across ads and landing pages. If a caption says 10% off, the landing page must say the percentage, not only the final price. Captions are text the shopper reads in the first seconds, so treat them like ad copy that has to match the page, not like decoration.
Mobile legibility is the other half of that advice. Short phrases of a few words read faster on a phone than a full sentence, which the caption phrasing design fields (max_words, max_chars) let you control per request.
A cost line for a 12-clip holiday set
Twelve clips at one caption job each are 12 x $0.20 = $2.40. If every clip is a silent loop, the only extra work is writing cues. If you also need a 15-second cut from a longer master, a video-trim job is $0.02, so 12 trimmed and captioned clips add $2.64 of post-production on top of the render cost.
Check the live price in GET /v1/catalog before committing the budget.
Sources
Related posts
More in Use cases
- Amazon main image bans promo text: where the Black Friday banner goes
Amazon's photography page says main images are pure white, fill 85-100%, with no promotional text. Put the sale banner on a video overlay. Read 2026-10-05.
- Amazon Sponsored Brands video at 20 s: two Omni 1080p clips, $3.86
Amazon allows 6 to 45 seconds, 16:9 only, and recommends 20 or less. At Sume rates, two 10-second Omni 1080p clips plus a Timeline stitch cost $3.86.
- Amazon "up to 50% off" needs 10% of ASINs at 50%: a caption check
Amazon says an up-to claim needs 10% of products at the maximum discount. A short Python check counts the ASINs and burns the matching caption cue.
- Animal Shelter Adoption Appeal as an AI Avatar Video: Cost
A shelter adoption appeal using an avatar and a scene photo: scripted scenes, per-scene seconds, request shape and Sume cost on each quality tier.
Written by Sume