# Video to Video AI > Restyle or recast an existing video with a text prompt. Keep camera and timing, switch models, quote credits first, and download a private MP4 on MixVio. - [Canonical Video to Video page](https://mixvio.ai/tools/video-to-video) ## Tool summary - **Media:** video - **Input:** One source video plus a text prompt - **Output:** Private MP4 whose length follows the source video - **Credits:** From 51 credits - **Limit:** Source 3–15s on the default route · optional look image ## How to use 1. **Upload the source video:** Start with a take you own or are authorized to edit. Generate stays disabled until a source video is attached. Add an optional look image if you have a face, outfit, or place to match. 2. **Describe the change:** Write what should change and what must stay — wardrobe, subject, or setting. Example: replace the cat with a dog; keep camera, background, and timing. 3. **Generate and download:** Confirm the credit quote, generate a private MP4, compare it with the source video, then download or run another pass on the same take. ## Common uses - Keep the dance, restage the dancer - Change wardrobe or product on a locked video - Restyle a setting or recast a subject ## FAQ ### What is Video to Video AI? Video to Video restyles or recasts an owned video from a prompt. Change wardrobe, the subject, or the setting; camera and timing stay on the take unless you ask otherwise. MixVio returns a private MP4 whose length follows that video. It is not Image to Video, not Motion Control, and not a style-filter pack. ### How do I restyle a video with AI? Three steps in the generator on this page. 1) Upload the source video: a take you own or are authorized to edit. Generate stays disabled until a source video is attached. Add an optional look image if you have a face, outfit, or place to match. 2) Describe the change: write what should change and what must stay—wardrobe, subject, or setting. Example: replace the cat with a dog; keep camera, background, and timing. 3) Generate and download: confirm the credit quote, generate a private MP4, compare it with the source, then download or run another pass on the same take. ### Is MixVio Video to Video free? You need a MixVio account. New accounts receive 40 welcome credits valid for 7 days. Video to Video is available on the Free plan when you have enough credits for the quoted route. Failed runs cost no credits. MixVio does not offer a no-account or no-watermark trial. ### How do credits work? The default Kling 3.0 Omni route is billed by source length: 17 credits per second at 720p and 23 credits per second at 1080p. A 3-second 720p run is 51 credits. Wan 3.0 and Seedance 2.5 quote separately in the generator. Confirm the exact amount before you generate. Failed or cancelled tasks release the reservation. ### What kinds of prompts work best? Name one change and lock the rest. Keep camera, timing, and identity explicit, then request wardrobe, subject, or setting. Example: replace the cat with a dog; keep camera, background, lighting, and timing. Avoid a new shot, a longer video, motion that is not in the source, or a different look on each beat—this page applies one instruction to the whole take. ### Will the result match the source video exactly? No. Instruction edit can shift pixels outside the requested change. There is no brush, mask, timeline trim, video extend, or lip-sync on this page. Always inspect the private export before you publish. ### What if I am not satisfied with the result? Adjust the keep-and-change instruction, switch model, or replace the source video, then generate again. Each new run shows its credit quote before submit. Failed runs cost no credits. ### Which models can I use for Video to Video? This page switches Kling 3.0 Omni, Wan 3.0, and Seedance 2.5. Start with Omni for a 3–15 second instruction edit, Wan 3.0 for a 2–15 second whole-video restyle, and Seedance 2.5 for a 4–30 second owned take. Compare options in the Models section above. ### What input limits apply? Upload one owned MP4 or MOV. Add an optional JPEG or PNG look image up to 50 MB. Default Omni accepts 3–15 seconds and videos up to 200 MB; Wan 3.0 accepts 2–15 seconds and Seedance 2.5 accepts 4–30 seconds. Duration, resolution, and audio stay on the route defaults — they are not first-screen controls. ### Can I use Video to Video outputs commercially? Paid-plan outputs may be used commercially, subject to the Terms of Service and the current model terms. You must have rights to the source video, faces, brands, and any look image. Free-plan outputs are for evaluation and are not for commercial publishing. ### Are uploads and results private? Yes. Inputs and outputs remain private workspace assets. Signed URLs, asset IDs, and private prompts are not added to public page URLs or analytics. ### Should I use Video to Video or Image to Video? Use Video to Video when you already have a video and want an instruction restyle or recast. Use Image to Video when you start from an image and need invented camera motion. [Open Image to Video →](https://mixvio.ai/tools/image-to-video) ### Should I use Video to Video or Motion Control? Use Video to Video to restyle or recast the video itself. Use Motion Control when a performance should move a character photo — the image moves; the video is the motion source. [Open Motion Control →](https://mixvio.ai/tools/motion-control)