wan2.7-videoedit camera replication vs Sume's video_url edit
Alibaba names wan2.7-videoedit for effect and camera-movement replication. On Sume, video edit is the video_url field on the Gemini Omni Flash 1.1 id.

Alibaba recommends wan2.7-videoedit when you want effect replication or camera movement replication, at 720P or 1080P and 2-10 s. Sume has no model named for those two tasks. Its video edit is the video_url field on gemini-omni-flash-1.1, driven by a text prompt, with no field that takes a second clip as an effect or camera reference.
Alibaba's side is from its video generation page, read 2026-10-01. Sume's is from the Video Router docs.
What does Alibaba say wan2.7-videoedit is for?
Its table row reads: video editing, effect replication, camera movement replication, 720P and 1080P, 2-10 s. Separately, the page recommends happyhorse-1.0-video-edit for plain text-instruction edits such as style transfer and element replacement, and points to wan2.7-videoedit for the replication tasks.
How does Sume take a video to edit?
In POST /v1/video-router/generate, send video_url with a prompt that describes the edit and optionally resolution (default 720p). The docs state that video_url is the edit source, not a reference: it cannot be combined with image_url, end_image_url or reference_*_urls.
Use reference_video_urls when a clip should condition a fresh generation instead of being edited. The docs describe it as conditioning a new generation, not as an edit or a replication feature.
curl -X POST https://api.sume.com/v1/video-router/generate \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: video-edit-001" \
-d '{
"model": "gemini-omni-flash-1.1",
"prompt": "Replace the bottle with an apple. Keep everything else the same.",
"video_url": "https://example.com/clip.mp4",
"resolution": "720p",
"mode": "async"
}'Which fields differ for an edit request?
| Field | Rule on an edit |
|---|---|
video_url | The source clip; edit mode is triggered by it |
aspect_ratio | 400 if given; output follows the source clip |
duration | Not sent to the provider; only a reserve-estimate hint (default 8s) |
generate_audio: false | 400; native audio is always produced |
resolution | Optional; default 720p |
Is there another model that takes video_url?
Only one: the Genjutsu row takes video_url, as a Motion Transfer source, not as an edit. See Genjutsu is Motion Transfer only. For a comparable field-level story on another vendor, see Luma's edit and aspect ratio.
Sources
Related posts
More in Models
- Wan 3.0 Prime vs Wan 3.0: one wan-3.0 id on Sume
QwenCloud and Runway list Wan 3.0 and a Prime variant. Sume's catalog lists one id, wan-3.0, with 2-30 s clips. Check the catalog for limits.
- Wan 3.0 is public after invite-only: check Sume's wan-3.0 id
QwenCloud says Wan 3.0 went from invite-only on August 6 to public. On Sume, wan-3.0 is a catalog id; check the live catalog for limits.
- Wan-Animate-2 14B: Apache 2.0 weights vs Sume Motion Transfer
Wan-Animate-2 14B weights are Apache 2.0 and ship inference scripts. Sume's Motion Transfer takes one source video plus 1-8 reference images, 4-30 s.
- Wan-Animate-2 Base vs Distillation weights: which to download
Wan-Animate-2 ships Base and Distillation checkpoints, each with its own YAML config. What the card says, and the hosted limits Sume lists for comparison.
Written by Sume