Available now

Wan 2.7: video from references

By Alibaba

Wan 2.7 makes 2 to 15 second clips at 720p or 1080p from a prompt, a first and last frame, reference images or a clip to continue, and edits clips you upload.

Model ID
Maker
Alibaba

Input

Fields marked * are required. Review the price before generating.

Text prompt describing the desired video. Max 5000 characters.

Example prompt. Edit it or write your own.

Reference image for character/object appearance. Pass multiple images for multi-subject generation. Max 20 MB each.

Uploading a file needs an account. Create account

JPEG, PNG, WebP · up to 25 MB

Output video resolution tier.

Output video duration in seconds (2-10).

Additional settings

Content to avoid in the video. Max 500 characters.

When true, enables intelligent multi-shot segmentation. When false (default), generates a single continuous shot.

Random seed for reproducibility (0-2147483647).

Estimated total · 5 s, 1080p$0.80

You sign up first. Your settings are kept.

About this model

Overview

Wan 2.7 renders 2 to 15 seconds in 16:9, 9:16, 1:1, 4:3 or 3:4. Upload a WAV or MP3 to drive the clip; without one, text clips get background music made to match.

The image mode starts from a first frame, can end on a last frame, or continues a video clip of 2 to 10 seconds instead. The reference mode builds a 2 to 10 second scene around the people and objects in your images; the edit mode restyles or changes a clip of 2 to 10 seconds and can keep its sound. A negative prompt keeps things out.

What you can build

  • Clips cut to your own song or voice-over
  • A second part for a clip you already have
  • A move from a first frame to a last frame you choose
  • Short scenes with matching background music
  • A clip you already have, restyled or changed by a written instruction

Prompting tips

  • Upload your audio when the motion should follow its beat or speech; keep it within 30 seconds.
  • To continue a clip, describe what happens next, not the scene you already have.
  • Turn prompt rewriting off when the model must follow your exact words.
  • In the edit mode, add a reference image to show the look or object you want, and choose whether the clip keeps its own sound.

What to check

Check motion continuity, the ending transition and sound alignment across the clip.

Inputs

  • Prompt (required): Text prompt describing the desired video. Max 5000 characters.
  • Reference Images (optional): Reference image URLs for character/object appearance. Pass multiple images for multi-subject generation. Max 20 MB each.
  • Negative Prompt (optional): Content to avoid in the video. Max 500 characters.
  • Aspect Ratio (optional): Aspect ratio of the generated video. Options: 16:9, 9:16, 1:1, 4:3, 3:4.
  • Resolution (optional): Output video resolution tier. Options: 720p, 1080p.
  • Duration (optional): Output video duration in seconds (2-10). Options: 2, 3, 4, 5, 6, 7, 8, 9, 10.
  • Multi Shots (optional): When true, enables intelligent multi-shot segmentation. When false (default), generates a single continuous shot.
  • Seed (optional): Random seed for reproducibility (0-2147483647).

Output and limits

  • Length: 2 seconds to 10 seconds.
  • Resolutions: 720p, 1080p.
  • Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4.
  • Readable text inside the video, especially small text, is unreliable.
  • Scenes with several interacting characters can lose consistency.
  • Longer clips can show more motion drift or scene changes.
  • Low-quality inputs cause glitches: heavy compression, collages, screenshots and text overlays.
  • Some prompts or outputs can be refused by safety checks.

Price guide

  • 1080p / 720p: $0.1585 / second
  • The total for your settings is rounded up to the next cent.