Edits driven by one instruction
The prompt is required and does the main work. Describe the change in plain language, such as a new outfit color, a different art style or a changed background, and the model applies it across the clip.
Wan 2.7
Edit a video with a single prompt
Wan 2.7 Video Edit restyles or modifies an existing clip from a text instruction, optionally guided by a reference image for outfits, characters or style.
Try Video EditFAQ
Yes. Every run needs a text instruction and a source video. The reference image is optional.
Changes you can describe clearly: clothing, colors, visual style, characters and setting. Keep each run focused on one or two changes, and say what must stay the same.
Add one when the exact look matters, for example a specific product, garment, character or style frame. Only 1 image is accepted per run.
It starts at 264 credits and the price depends on resolution and duration. You see the exact cost before generating. Failed generations are refunded automatically. New accounts get 50 free credits, and the Free plan adds 50 credits monthly.
Content you generate can be used in ads, stores and social media, subject to NeoMundo's terms of service and the model provider's policies. You should own or have permission to use the source footage and any reference image.
No. It is a generation model that edits one clip per run from a prompt. Cutting, trimming and sequencing still happen in your usual editing app; use this model for the changes that are hard to do by hand.
Video: MP4, MOV or WebM up to 100 MB. Reference image: JPG, PNG or WebP up to 10 MB. Output resolution is 720p or 1080p.
Upload a video, describe the change and optionally add a reference image. The credit cost is shown before you start, and new accounts get 50 free credits.
Try Video EditThe prompt is required and does the main work. Describe the change in plain language, such as a new outfit color, a different art style or a changed background, and the model applies it across the clip.
Add up to 1 image when words are not enough: a specific garment for an outfit change, a character design, or a style frame you want the whole clip to follow.
The source video stays the base of the result. You are editing an existing shot rather than generating a new one from scratch, so composition and action come from what you filmed.
Choose 720p for drafts and quick social posts, or 1080p when the edited clip goes into an ad, a product page or a longer edit.
The model starts at 264 credits. The exact cost for your clip and resolution is shown before you generate, and failed runs are refunded automatically.
Wan 2.7 video edit is a prompt-based AI video editor from the Wan model family by Alibaba. Instead of masking, keying or rotoscoping frame by frame, you give it a clip and a text instruction, and it restyles or modifies that clip. On NeoMundo it is listed as Video Edit in the Motion & swap studio.
The model is useful when the shot itself is right but something in it is not: the outfit is the wrong color, the look does not match your brand, or the scene needs a different visual style. You edit video with a prompt, and when a written description is too vague you add a reference image so the model can see exactly which jacket, character or style you mean.
Inputs: 1 source video (required; MP4, MOV or WebM up to 100 MB), a text prompt (required) and 0 or 1 reference image (JPG, PNG or WebP up to 10 MB). Output: 720p or 1080p. Pricing starts at 264 credits and depends on resolution and duration. Videos usually finish in about 1 to 5 minutes.
How it compares with its neighbours: Kling 3.0 Omni Video Swap is built to replace a person or product with up to 4 subject images, and Wan 2.2 Animate replaces a character with no prompt at all. Choose Video Edit when the change is best described in words, such as style, clothing, color or setting.
Use a clear, steady video where the area you want to edit is visible and reasonably large in frame. Short, focused clips are easier to edit than long sequences with many cuts.
State what to change and what to keep, for example: "Change her jacket to a black leather jacket, keep her face, the room and the camera move." One or two changes per run give more predictable results than a long list.
For a specific product, garment, character or art style, upload one sharp reference image and mention it in the prompt, such as "use the outfit from the reference image".
Check the credit cost, generate, and review the result. Refine the prompt and run again if needed, then download the version you want to keep.
Show the same model in another color or garment from your collection by adding the product photo as the reference image.
Give stock or phone footage a consistent look, such as an illustrated, film or animated style, so a campaign feels like one series.
Turn a performer into a costumed or stylized character using a character design as the reference, while the action stays the same.
Refresh an older clip for a new season or promotion without booking a new shoot, then post it on short-form channels.