Create variations from a reference video

AI Video to Video Generator

Video-to-video uses a reference clip and a prompt to guide a new AI video. Describe the changes you want, then choose an available model and output settings.

Video to video

Reference-guided video

Prompt-guided remix

Reference audio

Select Image

Select Video

280 credits

Get Inspired

Hover to autoplay the gallery, recreate the setup instantly, or open any clip in a larger player.

Video example seedance-2-laundry-balcony

Remix a reference video in three steps

Video

Choose a model that supports video references, describe your changes, and review the generated result.

Upload a reference video

Choose a model that accepts video references, then upload a clip with a clear subject and readable movement.

Describe the changes

State what you want to change and what you want to keep, such as keeping the camera direction while changing the scene mood.

Generate and review

Choose the settings available for your model, generate a video, then preview it and refine the prompt or reference.

Core features of AI Video to Video Generator

Use reference clips, prompts, and model-specific settings to explore new video variations.

Reference-video guidance

Use a source clip to guide a new AI video. The output may vary from the original timing, framing, and movement.

Prompt-guided transformation

Describe the subject, action, style, or atmosphere you want to change. Prompts guide generation rather than precise frame-by-frame editing.

Model-specific reference inputs

Available reference inputs depend on the selected model. Use video references and, where supported, image references to guide the result.

Compare available models

Switch between available models and review their outputs using your own reference material and prompts.

Model-specific output settings

Choose from the aspect ratios, durations, and resolutions available for the selected model. Options can differ between models.

Reuse inspired prompts and settings

Load an example to reuse its prompt and settings, then add your own reference material. Examples do not guarantee access to a source video or an identical result.

AI Image & Video Plans

Compare monthly and yearly plans for AI image and video generation.

LiteSAVE 30%

Perfect for individual creators

$13.9/ month

$166.8 USD billed yearly

48,000 credits per year, granted all at once
Image generation with GPT Image 2 / Nano Banana 2
Seedance 2.0 Fast video generation
Video output up to 720p / 1080p depending on model
Audio-video sync for avatar workflows
2 concurrent generation jobs
Normal generation speed
Private videos
All annual credits are granted at once after successful payment and expire at the end of the annual subscription period.

Popular

ProSAVE 30%

For professional content creators

$34.9/ month

$418.8 USD billed yearly

138,000 credits per year, granted all at once
Image generation with GPT Image 2 / Nano Banana 2
Seedance 2.0 Fast video generation
Video output up to 720p / 1080p depending on model
Audio-video sync for avatar workflows
3 concurrent generation jobs
Fast-track generation lane
Private videos
No watermark
Commercial usage rights
Priority access to newly added models
All annual credits are granted at once after successful payment and expire at the end of the annual subscription period.

UltraSAVE 30%

Ultimate power for businesses

$69.9/ month

$838.8 USD billed yearly

300,000 credits per year, granted all at once
Image generation with GPT Image 2 / Nano Banana 2
Seedance 2.0 Fast video generation
Video output up to 720p / 1080p depending on model
Audio-video sync for avatar workflows
4 concurrent generation jobs
Fastest generation lane
Private videos
No watermark
Commercial usage rights
Priority access to new model releases
Early invites to selected new features
Priority support
Lowest cost per 100 credits
All annual credits are granted at once after successful payment and expire at the end of the annual subscription period.

All payments are securely processed by Stripe.

Frequently Asked Questions

Common questions about reference videos, remix prompts, and getting better video-to-video results.

What kinds of clips work best for video-to-video generation?

Start with a clip that has a clear subject, readable movement, and stable framing. Choose a model that accepts video references and follow its upload requirements.

Do I still need a prompt if I upload a reference video?

Yes. Use the prompt to describe your intended changes and what you would like to keep. The reference guides generation, but neither input guarantees an exact match.

How do I get better video-to-video remixes?

Use a clear source clip and change one aspect at a time. Preview each output, then refine the prompt or reference before generating another variation.

Will the original motion and frames be preserved exactly?

No. The model uses the reference to generate a new video, so motion, framing, and details may change. This workflow does not guarantee precise frame-by-frame edits.