Home/ Video to Video AI Editor
Video to Video AI Editor
Keep the footage, change everything else. Upload a clip, describe the edit — wardrobe, weather, style — and WAN 2.7 Video Edit rewrites the picture while your timing survives.
Video-to-video edits a clip you already have instead of generating a scene from scratch. The source video keeps its motion and framing; your prompt decides what changes — a style transfer, a wardrobe swap, day to night. Creovox opens WAN 2.7 Video Edit on this page, which takes a source clip, lets you keep or replace the audio, and renders at 720p or 1080p from 150 credits.
- Model
- WAN 2.7 Video Edit
- Starts with
- A source video
- Output
- MP4
- Clip length
- 2–10 s
- Resolution
- 720P / 1080P
- Sound
- Optional, extra credits
- Credits
- from 150 credits
- Commercial use
- Included — see pricing
The motion is yours
Timing, blocking and camera work come from your footage. The model repaints, it does not re-choreograph — which is what makes the edit usable.
Point the prompt at the change
"Same shot, raining, night, neon reflections" edits what you name and leaves the rest alone. A narrow instruction beats a full scene rewrite.
Keep or regenerate the audio
An audio setting lets the original soundtrack ride along or be replaced — useful when the edit changes the scene but not the dialogue.
Reference images steer the look
Optional references push the output toward a specific grade, palette or character design when words alone undersell it.
How to edit a clip with a prompt
- Step 1Upload the sourceSteady footage with a clear subject edits cleanest. Heavy handheld shake carries through.
- Step 2Name one changeResist the urge to rewrite everything — pick the swap you actually need and generate.
- Step 3Set duration and audioMatch the source length or cap the output; keep the original audio or let it regenerate.
- Step 4Compare against the sourceIf the edit drifts from the footage, narrow the prompt and run again rather than adding more instructions.
What a prompt-driven edit is good for
Restyle finished footage
Push a live-action take toward anime, a film-stock look or a different grade without a reshoot.
Swap wardrobe or props
Change what a subject wears or holds while the performance stays identical.
Relight the day
Day to night, sun to rain — fix the conditions the shoot could not control.
Localize an ad
Same cut, different market: adjust the visible setting without editing the timeline.
Know before you generate
- A source video is required — for a scene from scratch, use text-to-video.
- Output runs in short segments, up to about 10 seconds; match-source keeps your clip’s length.
- 720p and 1080p output, from 150 credits; cost shows before you run it.
- To add seconds to a clip rather than restyle it, that is the video-extend page — a different model.
Video-to-video questions
How long can the edit be?
Fixed lengths up to about 10 seconds, or match the source clip’s duration.
Does it keep my audio?
You choose — an audio setting keeps the original track or regenerates it.
How much does it cost?
From 150 credits depending on output settings; the studio shows the price first.
What is the difference from video extend?
Video-to-video changes what a clip looks like. Extend adds new seconds to the end of it.
What source footage works best?
Clear subject, steady camera. The model preserves your motion, so chaotic footage stays chaotic.
Restyle a clip you already shot
Open WAN 2.7 Video Edit and point a prompt at the change.
Keep going
Nearby tools and pages people usually open next.
Other model pages
Each page opens its own model in the same studio.
Pricing

