Meta AI effects in Stories vs making your own AI video clip
Instagram's Meta AI Stories effects transform photos with one tap and are US-only. For an AI video clip instead, generate a 9:16 clip. Steps and limits.

Meta AI's new effects in Instagram Stories are photo effects: Instagram describes over 30 of them, powered by Muse Image, that "transform your photos with a single tap," and says they are "available in the US only." If you want a video clip for a Story instead, you generate one yourself as a 9:16 clip with a video model, which is a separate route.
The Instagram facts come from Instagram's own announcement, read 2026-09-29. The page does not say whether the effects apply to video, so this post does not claim they do. The generation facts come from the Sume video generation docs.
What are the Meta AI effects in Stories?
Per the announcement, they are one-tap effects that can "subtly enhance the lighting in your photos" or let you "try on new looks," built on Muse Image, described as the first AI image generation model from Meta Superintelligence Labs. They run inside Instagram, on the photo you pick, in the regions the page names.
| Meta AI effect in Stories | Your own AI video clip | |
|---|---|---|
| Input | A photo, per the announcement | A text prompt, optionally frames |
| Output | An effect on that photo | A new video, aspect_ratio 9:16 |
| Availability | US only, per the announcement | Sume API |
| Where it runs | Inside Instagram | Outside Instagram |
How do I make my own vertical clip for a Story?
Send POST /v1/videos with a model, a prompt, aspect_ratio set to 9:16, and a duration. The catalog lists seedance-2 with 9:16 among its aspect ratios and durations of 4 to 15 seconds, and gemini-omni-flash-1.1 accepts 3 to 10 seconds in 16:9 or 9:16 with native synced audio. Read supported_aspect_ratios and supported_durations for the model you choose before you submit.
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 vertical clip of a sneaker on a sunlit windowsill, slow push-in",
"aspect_ratio": "9:16",
"duration": 5
}'What if I want a still instead of a clip?
Then generate an image at a vertical ratio. The Image API lists 9:16 among the aspect ratios in its model examples. This gives you a new picture rather than an effect on your own photo, which is the difference from the Instagram feature. Sume's image endpoint takes a prompt and optional reference images, so you can also start from a product shot, but that is generation, not a one-tap restyle of a photo in the Stories camera.
Which route should I pick?
Use the in-app effect when you want to restyle a photo you already have, and you are in the region the announcement names. Generate your own clip when the Story needs motion or a scene that does not exist yet. For Story dimensions and ad placement rules, see Instagram and Facebook Stories ad specs.
Sources
Related posts
More in Use cases
- Microdrama AI video: keep the same characters across episodes
Sume's video API has no cast-lock switch. To keep a character across episodes, start every shot from the same still with first_frame and reuse references.
- Microdramas on YouTube: how Shorts series fit the format
YouTube says microdramas passed 6.5 billion views in early 2026 and Shorts series now supports episodes. What the format needs and how to render episodes.
- Nano Banana consistent characters and 14 objects: the Sume cap
Google says Nano Banana 2 keeps up to 5 characters and 14 objects consistent. Sume takes 10 references per request, so 19 subjects need a plan.
- Nano Banana character consistency: reference limits and a Sume request
Google documents character reference slots for its Nano Banana models. The limits, and how to keep a character across images with input_references on Sume.
Written by Sume