Meta's video from static assets is GA: prepare six photos first
Advantage+ can now generate video from existing images. Prepare the photos so the clip keeps your product intact, and price a backup run on Sume.

Meta said on Oct 6, 2026 that in Advantage+ creative, advertisers and agencies can now "generate video from existing static assets", and that this feature is generally available (Meta for Business, read 2026-10-07). The quality of that video depends on the photo you give it. Prepare six photos per product, and keep a pinned-frame backup on Sume: a 5-second clip at 720p on Gemini Omni Flash 1.1 costs $0.625.
Why the photo decides the clip
A still-to-video system animates what it sees. If the product is small in the frame, cropped, or partly covered by text, the motion tends to invent the rest. The fix is in the input, before any generation.
Six photos to prepare
Use this list for each product.
- One clean hero shot, product centered, label readable at full size.
- One in-hand or in-use shot that shows true scale.
- One close-up of the label or the logo, because text is what drifts first.
- One vertical 9:16 version of the hero, so the first frame already has the final shape.
- One flat-color background version, for a clean first or last frame.
- One lifestyle shot with space at the top or bottom for captions.
A pinned backup on Sume
Meta's tool is a good first try because the cost is inside your ad budget and the setup is a toggle. A pinned backup helps when the result changes the pack, the label or the colors, or when you want the same clip on a second platform.
On Sume, POST /v1/videos takes the photo in frame_images with frame_type: "first_frame" for image-to-video, as the Video Generation docs show. If you send both frame_images and input_references, frame_images decides the mode.
{
"model": "gemini-omni-flash-1.1",
"prompt": "Slow push-in on the bottle, soft window light, label stays sharp",
"duration": 5,
"resolution": "720p",
"aspect_ratio": "9:16",
"frame_images": [{"type": "image_url", "image_url": {"url": "https://example.com/hero-9x16.jpg"}, "frame_type": "first_frame"}]
}What six backups cost
Gemini Omni Flash 1.1 is $0.10 per second at 720p on the provider list, so Sume bills 5 x $0.10 x 1.25 = $0.625. Six photos at 5 seconds are $3.75. At 1080p the rate is $0.1875 per second, so one clip is $0.9375 and six are $5.625.
A pure sume/auto request is 8 seconds and 720p by default, which is $1.00. Name the duration if you want the lower price.
Check before you publish
Meta's tool keeps the work inside Meta. A file from Sume goes anywhere. Check each clip with the first and last frames before you spend media budget, and keep the original photo next to it for the compare.
Record the prompt with the clip. A short line such as "slow push-in, soft window light, label stays sharp" is enough, and reusing the exact text across six products keeps the set looking like one campaign.
Treat a text-heavy product as a special case. If the label has small print, shoot the close-up at the highest resolution you have and use 1080p for that one clip. The extra $0.9375 per 5-second clip is small next to a recall of a mislabeled ad.
Also name each file by SKU and angle, so a reviewer can match a clip with its photo in seconds. Finally, avoid photos with a model's face if you plan to dub the clip into another language, because lip movement in the source will not match new audio. Use product-only photos for localized runs.
Sources
Related posts
More in Use cases
- Microscope-style science clip with Gemini Omni Flash: prompt and cost
Google's own Omni 1.1 demo prompt is a diatom micrograph. Here is how to write one, draft it at 360p, and what 8 seconds costs on Sume at each resolution.
- Midday to golden hour in a video: Omni edit or first/last frame
Change the time of day in a clip you already have. When a Gemini Omni Flash edit fits, when first and last frames fit, and what a 6-second pass costs on Sume.
- Mobile pet groomer reel from one photo: Seedance 2.5 image-to-video
Turn one photo of a groomed dog into a vertical reel clip with Seedance 2.5 on Sume: first-frame input, 4 to 30 seconds, native audio, async job.
- Monthly investor update video with an AI avatar: 60 s per job
An avatar update longer than 60 seconds has to be split. How to cut a monthly investor or board update into three jobs, keep one presenter, and price it.
Written by Sume