Wan 3.0 web page to video: link input and how to do it on Sume
Wan 3.0 can read a public web link to make a video on Alibaba's API. Sume's wan-3.0 has no link field, so fetch the page yourself and send text and images.

Wan 3.0 accepts a public web link as input on Alibaba's API and fal's page, and turns the page into a video; Sume's wan-3.0 does not accept a link. To make a video from a web page on Sume, fetch the page yourself, keep the text and images you want, and send them as a prompt, references or the input of an agent run.
Alibaba's link rules are from its API reference; fal's Wan 3 page says web pages need thinking mode. Sume's side is from Video generation and Create a run. All read 2026-09-29.
What does Alibaba say about web page input?
| Item | Stated rule |
|---|---|
| Protocol | HTTP or HTTPS |
| Access | Publicly accessible pages that do not require login |
| Per request | One link, or one file; the two are mutually exclusive |
| With references | Can be combined with reference images, videos and audio |
| Prompt extension | prompt_extend must be true |
Does Sume's wan-3.0 take a URL?
Only as media. input_references and frame_images take HTTPS URLs to image, video or audio files, and Sume fetches them. A page URL is not a media file, and the catalog states that web_url is not exposed in v1.
How do I turn a page into a video on Sume?
- Fetch the page with your own code and keep the headline, the key claims and the images you have the right to use.
- For a single clip, write a prompt from those claims and pass your chosen image as
first_frame, onwan-3.0. - For a finished video, send an Agent Completion or a Format run with the text in
input. A page URL insideinputis allowed and, unlike an image or video URL, does not count toward the run's media budget; the docs do not say the agent opens it, so send the text too.
What should I check before I publish?
That the video says only what the page says, and that you may reuse the page's images and words. A model that reads a page can still compress a claim into something the page did not say, so review the clip against the source.
Sources
Related posts
More in Models
- Kazakh speech to text API: Nova-3 and Sume language_code
Deepgram added Kazakh to Nova-3 on Sep 3, 2026. Sume STT 1.0 takes an optional language_code hint, or auto-detect. How to send audio and check the output.
- FLUX video edit API: 720p downscale, and video_url edits on Sume
BFL's edit-a-video endpoint downscales sources above 720p. Sume lists no FLUX video; its edit path is video_url with gemini-omni-flash-1.1 only.
- Gemini 3.5 Transcribe API limits: 1 hour vs Sume STT's 10 minutes
Gemini 3.5 Transcribe takes up to 1 hour per request, 30 minutes with diarization or word timestamps. Sume STT takes 600 seconds; split longer audio.
- Gemini API speaker diarization: 8 speakers; Sume STT has no labels
Gemini 3.5 Transcribe diarizes up to 8 speakers, 3+ experimental, within 30 minutes per request. Sume STT returns word timings but no speaker labels.
Written by Sume