The output loses the source motion
Use a shorter, clearer source video and keep prompt changes limited.
MODEL
ADD SOURCE VIDEO
MP4, MOV · max 100MB · 3-15s
PROMPT
Aspect Ratio
Duration
Resolution
CREDITS REQUIRED
250 Credits
Video to Video demo video
Start a generation to see your latest videos here.
VIDEO TO VIDEO AI
Quick answer
Video to video transforms an existing clip into a new AI video. Upload the source video for timing and motion, then use the prompt to describe what should change.
Input requirements
Match the source type to a conservative first motion target. That improves the keeper rate before spending credits on longer clips.
| Input | Best for | Avoid |
|---|---|---|
| Source video | Motion, timing, camera rhythm, pose, or action reference | Long clips, rapid cuts, or heavy camera shake |
| Prompt | One clear transformation goal and preservation rules | Conflicting subject, style, action, and camera changes |
| Output settings | A supported 5, 10, or 15 second 720p or 1080p result | Assuming unsupported settings will be sent to the provider |
Failure reasons
Use a shorter, clearer source video and keep prompt changes limited.
Use one specific style direction and remove contradictory wording.
Ask for one transformation goal instead of rewriting the whole clip.
FAQ
It is best for remixes, style changes, subject changes, and motion-preserving transformations from existing footage.
Short clips between 3 and 30 seconds are easier to control and estimate.
No. This Wan 2.6 workflow submits the source video and prompt. Use Reference to Video when you want an image-guided workflow.
Related pages
Use character replacement when the main goal is swapping the visible subject.
Open replace characterREFERENCEUse a reference image when you want image guidance without a source video.
Open reference to videoPROMPTGenerate a new clip from text when no source footage is available.
Open text to video