Wan 2.7 Reference-to-Video
[Core Function] Wan 2.7 Reference-to-Video is a highly capable character/entity reference video model. [Strengths] It natively supports entity reference, voice customization, and playbook-based video generation from a single storyboard. [Best For] Highly recommended for: creating consistent video series, brand mascot animation, and storyboard-driven storytelling. [Limitations] Do NOT use this model for simple, single-image direct animation (use I2V instead). [Routing] Use this by default for complex character consistency and storyboard generation tasks on Alibaba.
Authorizations
API Key authentication. Format: Bearer YOUR_API_KEY.
Body
At least one of reference_images, reference_videos, or first_frame_image must be provided.
Video content description (supports Chinese and English, up to 5000 characters). Can reference media using 'Image1/Image2' or 'Video1/Video2' format
1 - 5000"The character walks through a blooming garden"
Reference image URLs. At least one media parameter (reference_images, reference_videos, or first_frame_image) must be provided. Combined with reference_videos, max 5 total
Reference video URLs. At least one media parameter (reference_images, reference_videos, or first_frame_image) must be provided. Combined with reference_images, max 5 total
First frame image URL. At least one media parameter (reference_images, reference_videos, or first_frame_image) must be provided. Max 1, not counted in the 5-item limit
"https://example.com/first_frame.jpg"
Negative prompt describing unwanted content
500"blurry, distorted, low quality"
Video resolution tier
720P, 1080P "1080P"
Video aspect ratio (auto-ignored when first_frame_image is present)
16:9, 9:16, 1:1, 4:3, 3:4 "16:9"
Video duration in seconds. Dynamic range: [2,10] when reference_videos is present, [2,15] when only reference_images
2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 5
Enable intelligent prompt rewriting
true
Random seed for reproducibility
0 <= x <= 214748364742