Wan 2.7 Video and Image in One Line
Wan 2.7 covers instruction-based video editing, reference and image driven generation, and 2.7 text-to-image and image edit in a single model family.
- Alibaba
- 8 variants
- From $0.03 per image
- Price
- See price
- Speed
- ~2 min
- Best for
- Video to Video
- Price
- See price
- Speed
- ~8 min
- Best for
- Video to Video
- Price
- See price
- Speed
- ~3 min
- Best for
- Text to Video
- Price
- See price
- Speed
- ~1 min
- Best for
- Image to Video
- Price
- See price
- Speed
- ~10s
- Best for
- Text to Image
- Price
- See price
- Speed
- ~10s
- Best for
- Image edit · Image to Image
- Price
- See price
- Speed
- ~25s
- Best for
- Text to Image
- Price
- See price
- Speed
- ~25s
- Best for
- Image edit · Image to Image
More from Alibaba.
A family groups the versions of one model line. Open one to compare its variants side by side.
Working with Wan 2.7.
Wan 2.7 is the broadest Wan release on each::labs: it spans video editing with style transfer and reference-image edits at 720P/1080P, text, image and reference video generation with optional audio and auto background music, and a text-to-image and image edit pair with improved prompt fidelity.
When to choose Wan 2.7
Pick Wan 2.7 when one family has to cover both stills and video — the same prompt conventions carry across both. Wan 3.0 is newer for pure video generation; Wan 2.2 is the specialist for transferring motion onto existing footage.
Choosing between the variants
Video Edit takes existing footage and preserves or regenerates its audio. Reference to Video holds a character across single or multi-shot scenes. Text to Video and Image to Video generate from scratch with prompt enhancement, while Text to Image and Image Edit handle stills.
Questions about Wan 2.7
The short answers. For anything specific to your setup, talk to our team.
Yes. The Video Edit variant can either preserve the source audio or regenerate it, so a restyled clip does not have to lose its original sound bed.
The model expands a short prompt into a fuller scene description before generating. It helps thin prompts, and can be left off when you have written the shot precisely.
The Reference to Video variant does — it supports single or multi-shot output while keeping character and object appearance consistent from a reference image.