Amazon Sponsored Products video: specs and guidelines
Sponsored Products video takes 1–5 silent feature videos per product: 7 seconds or longer, 16:9 (1920×1080 recommended), MP4 or MOV up to 500 MB.

Amazon Sponsored Products video lets you add 1–5 product feature videos per product to a Sponsored Products ad group. Each video is 7 seconds or longer, horizontal 16:9 (1920×1080 recommended; vertical 9:16 is due in 2026), an MP4 or MOV of up to 500 MB, and silent, because Amazon removes any audio. Each one shows a feature of the product, which should fill at least half the frame, ideally with a person using it; slideshows, letterboxing, logos that aren't on the product, emojis, and talking-head videos are not allowed.
Amazon's rules are quoted from its Sponsored Products video guide, read on 2026-09-28; Amazon can change them. Sume facts come from the Video generation and Timeline 1.0 docs. Sponsored Brands video, which keeps its sound, is covered in Amazon Sponsored Brands video specs.
What are the Sponsored Products video specs?
Amazon may clip a long video to the part most relevant to the shopper; its example is a two-minute upload shown as 30 seconds.
| Spec | Amazon's rule |
|---|---|
| Videos per product | 1 to 5 in an ad group |
| Length | 7 seconds and longer, no maximum |
| Layout | Horizontal 16:9; vertical 9:16 to be available in 2026 |
| Resolution | 1920×1080 recommended |
| Formats | MP4 or MOV |
| Audio | None; audio is removed |
| File size | 500 MB or less |
| Feature title | Required; up to 3 words, 15 characters or less |
| Description | Optional; up to 6 words, 55 characters or less |
What can a product feature video show?
One feature, benefit, or use of a single product, not a brand story or a collection. Amazon's moderation team reviews every new video against these rules:
- The product takes up 50% or more of the frame, ideally with a person using it.
- No blurry frames, slideshows, or letterboxing; Amazon's don'ts also list on-screen captions.
- No explicit branding or logos unless they are on the product, no emojis or emoticons, and no nudity.
- No talking-head videos that need a script or caption to explain the product.
- The feature named in the title must be shown in the video, or the video is rejected.
How many feature videos should each product get?
Up to five per product in an ad group, and Amazon recommends more than one: a product with a single feature video doesn't get the interactive thumbnails, and up to three thumbnails show at a time. Amazon's FAQ adds that you will be able to upload one video per child product, such as a color or size, so the video matches the variation on the product detail page.
Can I make a Sponsored Products video with AI?
Only with care. Amazon's guide asks for real products demonstrating real usage and warns that overly animated videos or renders may erode shoppers' trust. A generated clip fits only if it shows your real product exactly as it is, doing only what it really does, and Amazon still moderates it.
With Sume, start each clip from a product photo on POST /v1/videos: the photo as the first frame (frame_images with frame_type first_frame), aspect_ratio 16:9, and generate_audio: false, since Amazon strips audio anyway. The photo must be reachable over public HTTPS. Make one clip per feature; product photo to video covers the request and AI product demo videos covers showing one real action.
curl -X POST https://api.sume.com/v1/videos \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedance-2",
"prompt": "A hand lifts the bottle from a kitchen counter and opens the lid; the bottle fills most of the frame",
"frame_images": [
{
"type": "image_url",
"image_url": { "url": "https://example.com/bottle-packshot.png" },
"frame_type": "first_frame"
}
],
"aspect_ratio": "16:9",
"resolution": "1080p",
"duration": 8,
"generate_audio": false
}'How do I get an exact 1920×1080 file?
seedance-2 lists resolution tiers, not pixel sizes (supported_sizes is null), so conform the clip with POST /v1/timeline-1.0/render: output 1920 × 1080, and audio.mode "silence" with audio.duration_seconds from 7 up to the length of the clip, which sets the length without a sound file. Timeline reads only your workspace's media.sume.com files, such as that generated clip (which URLs each endpoint takes). Today a silent render still carries an audio track, which Amazon removes.
Sources
Related posts
More in Media tools
- Can a JPG be transparent? No: make a transparent PNG instead
No. A JPG has no alpha channel, so every pixel is opaque. Cut the subject out and save it as a PNG, a format that can store transparent pixels.
- Cinematic bars size: bar height for 2.39:1 and 2.35:1
Cinematic bars on a 1920×1080 video are 138 px each for 2.39:1 and 131 px for 2.35:1. The formula, sizes for 720p to 4K, and how to add them.
- How to combine photos and videos into one video
Combine photos and videos into one video on a single timeline: each photo held for a set time, each clip cut to length, one soundtrack under it all.
- Convert 4:3 to 16:9 video: pillarbox, crop, or stretch
Convert 4:3 video to 16:9 three ways: pillarbox it, crop 12.5% off the top and bottom, or stretch it. With FFmpeg and the Sume API.
Written by Sume