WAND-Vega for e-commerce video: what Tencent says, what Sume lists

Yes, Sume can list WAND-Vega for e-commerce video, but only as text-to-video, and only where the Tencent switch serves it. Tencent's WAND page describes WAND-Vega as a content generation model trained for vertical industries, with e-commerce and comic-style short drama named as scenarios, covering image and video generation. On Sume the video side appears as three rows: wand-vega-1.0-lite, wand-vega-1.0-pro and wand-vega-1.5-pro, each 4 to 15 seconds with six aspect ratios and a generate_audio option.

What do the three rows accept?

The Video Router doc marks all three as Tencent VOD only and text-to-video only. That means no first frame, no reference images and no source clip: a product shot cannot be pinned to your own photo on these rows.

When would I pick it?

Pick a WAND-Vega row when a product ad can start from a written brief and you want a high-resolution output option. If the ad must show your exact packaging, use a row with a first frame or reference images instead (wan-3.0, seedance-2.5, gemini-omni-flash-1.1). The Tencent page says nothing on price or API detail, and this post does not claim any benchmark; read the Sume catalog row for the rate.

Why might the row be missing from my list?

The doc says the Tencent rows are in the list "only where the Tencent switch serves them". An id that is documented but not in your GET /v1/videos/models answer returns 404 model_not_found on submit. Check the live list in the environment you call, not the docs alone.