Drag and drop media files from your computer, paste from clipboard (Ctrl/Cmd+V), or provide a URL. Accepted: .jpg, .jpeg, .png, .webp
Ready to Generate
Configure your inputs and click run to generate an image preview.
Seedance 2.0 Mini is a high-speed, high-efficiency image-to-video (reference image/video generation) model provided by BytePlus. It inherits the outstanding visual expressiveness of its predecessor model while achieving dramatic improvements in light-weighting and inference speed. By optimizing resources, it outputs videos with surprisingly smooth movements and realistic textures.
It faithfully reproduces the details of characters and backgrounds in still images (reference images), while creating natural camera work and dynamic changes that follow the laws of physics. It supports a variety of styles, including cinematic live-action, 3D, and surrealism, and can flexibly generate videos from 4 seconds up to a maximum of 15 seconds. It is also designed to be suitable for audio synchronization (lip-sync) and adding text.
Ideal for creative scenarios that require speed and trial-and-error, such as prototype development and rapid content creation for SNS, it instantly elevates any user's inspiration into high-quality visuals.
| Resolution | Credits Consumed |
|---|---|
| 480p(credits/s) | 2 |
| 720p(credits/s) | 4 |
| Parameter | Specification |
|---|---|
| Core Capability | Reference image to video |
| Resolution | 480p,720p |
| Aspect Ratio | 16:9,4:3,1:1,3:4,9:16,21:9 |
| Duration | 4,5,6,7,8,9,10,11,12,13,14,15 |

BytePlus
An innovative multimodal generative model that creates cinematic videos up to 30 seconds long from reference images.

WAN
Next-generation model that transforms still images and audio into movie-quality videos with overwhelming generation speed

Google's fast multimodal video generation and editing model, which highly integrates audio and video

WAN
A next-generation integrated AI video generation model that achieves cinematic-level image quality and high consistency

Google's top-tier video generation model, crafting cinematic-level visuals and sound.

Kling
A movie-grade all-purpose video generation model that reproduces consistent characters based on images