Wan 3.0 for E-commerce: Product Demos and Video Ads

Key Takeaways

• This is not the short-form storytelling guide — it's the e-commerce angle: turning product photos and footage into demo clips and ad creative.
• Wan 3.0's i2v endpoint takes a product photo straight to video, with end_image available for before/after or unboxing framing.
• ref2v accepts up to 10 images, 5 videos, and 5 audio files, useful for keeping multiple product angles consistent in one generation.

Wan 3.0 for Short-Form Video: Scene and Structure covers narrative and camera-structure technique for storytelling clips — pacing, shot sequencing, and using aspect_ratio/duration for creative video. This piece is narrower and more commercial: turning product photos and footage into demo videos and ad creative for online stores. For the full request-schema rundown across all three variants, see the Wan 3.0 overview.

What Wan 3.0 Actually Offers for Product Content

Three endpoints are live on Siray: alibaba/wan-3.0-t2v, alibaba/wan-3.0-i2v, and alibaba/wan-3.0-ref2v. Output resolution is 480p, 720p, or 1080p (no 4K), duration is any integer from 2 to 30 seconds, and aspect_ratio covers six values: adaptive, 16:9, 9:16, 1:1, 4:3, and 3:4. audio_enable is a boolean with no default declared in the spec, so it needs to be set explicitly rather than assumed. None of the three variants exposes a negative_prompt field. There is no dedicated "product mode" or e-commerce-specific parameter in the schema — the fields above are the whole surface, and a workflow built around a flag that doesn't exist will fail at request time.

Turning a Product Photo Into a Demo Clip (i2v)

i2v takes a required image field and an optional end_image, setting the first and last frame of the generation. For a product demo, that means a static product photo as the start frame and a packaged/in-use shot as the end frame, with Wan 3.0 filling the motion between — closer to a directed transition than free-form animation. aspect_ratio matters for where the clip runs: 9:16 for Reels, TikTok, and Stories; 1:1/4:3 for feed and marketplace listings; 16:9 for a product page or YouTube embed. Because size and aspect_ratio are separate fields, the same shot can be re-rendered at a different ratio without re-describing the scene.

Multi-Angle Consistency With ref2v

Catalog work usually means more than one shot of the same product. ref2v accepts up to 10 reference images, 5 videos, and 5 audio files, letting several product angles or colorways inform a single generation rather than treating each angle as an unrelated prompt. This is a real schema capability, not a "catalog mode" — the same multi-input field used for any reference-driven generation, applied to product photography.

Batching a Catalog Without Overpromising Throughput

Running a full catalog through Wan 3.0 means issuing many requests, and that's a client-side concern: an async or thread-pool pattern on the caller's side, queuing items and handling retries, is the standard way to move through a batch. This article makes no claim, in either direction, about how many requests the platform itself can process at once — scaling guidance here is limited to how the calling code is structured. Per-second pricing for the plain Wan 3.0 endpoints isn't listed here; check the current rate on the model detail page in the Siray console before budgeting a run, rather than assuming it matches a different Wan 3.0 variant.

Developer Note

t2v requires model, prompt, duration, size, and aspect_ratio. i2v requires image plus an optional end_image. ref2v accepts up to 10 images, 5 videos, and 5 audios. All three share one Siray API key alongside every other model on the platform, so a pipeline that also needs an image model or a different video model for other placements is a config change, not a new integration.

FAQ

Can Wan 3.0 turn a single product photo into a video ad?

Yes, through i2v. The photo is passed as the required image field, and an optional end_image can set a second frame for a before/after or unboxing-style transition.

What's the longest a Wan 3.0 product video can be?

Up to 30 seconds. duration accepts any integer from 2 to 30 across all three variants — there is no separate "ad length" parameter.

Does Wan 3.0 have a dedicated e-commerce or product mode?

No. The request schema has no such flag; product use is achieved with the same t2v, i2v, and ref2v fields available for any other subject.

Create your free Siray account and start generating Wan 3.0 product videos.