AISVIT / AI Video / Video to Video

Runway Gen-4 Aleph Video to Video Editor

Edit video online with Runway Gen-4 Aleph. Transform scenes, background, objects, lighting, and style in a video-to-video workflow while keeping source motion.

About this model

Specialized video-to-video model for editing an existing clip with a text instruction: you can change scenes, add or remove objects, shift style, lighting, and atmosphere without rebuilding the shot from scratch.

When is this model useful?

Runway Gen-4 Aleph is strongest when you already have footage and want controlled transformation, not when you want to create a brand-new scene from nothing.

Best fit tasks

  • Video-to-video editing for ads, fashion clips, product videos, and social content when you need a new version from the same source footage.
  • Adding, removing, or replacing objects in frame, such as products, props, decor, crowds, visual effects, or distracting elements.
  • Changing location, season, time of day, weather, or overall atmosphere without rebuilding the full shot in a separate edit workflow.
  • Restyling a clip for a different brand look, campaign mood, or creative direction with the help of a prompt and an optional reference image.
  • Editing workflows where motion, pacing, and shot structure should stay recognizable while the visual content changes.

Main advantages

  • It is built for in-context editing, which means it treats your uploaded clip as the base and transforms it instead of inventing a completely unrelated shot.
  • It tends to preserve perspective, lighting, shadows, and frame-to-frame continuity better than broad remix workflows.
  • One model can handle several edit types in the same family: object addition, object removal, object replacement, relighting, environment change, and style transfer.
  • An optional reference image helps pull the result toward a target look, material, color language, or specific visual motif without complex manual compositing.

Limitations to know

  • In AISVIT, this is a short workflow: the model works with a 5 second clip segment even if the uploaded source video is longer.
  • Result quality depends heavily on the source footage. Noisy video, cluttered scenes, heavy occlusion, or complex interactions can reduce edit quality.
  • This is not the best tool for frame-perfect editing, tiny on-screen text control, or highly deterministic graphic placement.
  • It works better when each render has one clear edit goal. If you ask for a full scene rewrite, character change, style shift, and object swap all at once, control usually drops.

How to use this model

The best Aleph workflow is to start with a short source clip, describe one primary change clearly, and use a reference image only when it meaningfully improves style or object control.

Simple workflow

  1. Upload the source video you want to edit. Think of it as the motion and timing base that the model will try to preserve.
  2. Write the prompt in plain language and focus on what should change. For example: make it winter, replace the decor, remove a distracting object, change the lighting, or turn the scene into a different style.
  3. Choose the output aspect ratio. Available options are 16:9, 9:16, 4:3, 3:4, 1:1, and 21:9 depending on whether the clip is for a website, YouTube, Shorts, Reels, or a wider cinematic placement.
  4. Add a reference image if you want stronger control over the style, texture, color language, or appearance of a specific object or look.
  5. Use the seed value only when you want a more repeatable variation from similar settings.

Supported inputs

  • Required: one source video to edit.
  • In AISVIT, the input video must be under 16 MB, and this workflow uses only the first 5 seconds of the uploaded clip.
  • Optional: one reference image for style or content guidance.
  • In the AISVIT upload flow, common video formats such as MP4 or MOV are the most typical choice, and JPG, PNG, or WEBP are the safest image formats for the reference.
  • Available aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1, and 21:9.

What you get

  • A generated MP4 video file.
  • An edited clip that tries to preserve source motion, perspective, and scene logic while changing the elements described in the prompt.
  • A short video-to-video result oriented around roughly 5 seconds of processed footage in this integration.
  • Output framed in the selected aspect ratio: 16:9, 9:16, 4:3, 3:4, 1:1, or 21:9.

More Video to Video models

AISVIT pricing details

  • Runway Gen-4 Aleph: 18 credits per second
  • Typical 5 second run in AISVIT: 90 credits