Wan 2.7 - Image to Video

Authentication

authorization string required

All APIs require authentication via Bearer Token.

Authorization: Bearer YOUR_API_KEY

Parameters

model string required

Model ID to use for the request.

Value: wan27_i2v

input array required

The Vtrix generation API requires input to be an array. Model parameters must be placed under input[0].params. The cURL example is the request format for Vtrix customers, not the raw Alibaba DashScope request format.

input[0].params object required

Provider request body forwarded to Alibaba DashScope. Put request fields under input and generation options under parameters.

input object required

DashScope video request input.

input.prompt string

Text prompt describing the expected video content or edit instruction.

input.negative_prompt string

Negative prompt describing what should not appear in the result.

input.media array

Input media array. For image-to-video, use media types such as first_frame, last_frame, driving_audio, or first_clip according to the generation mode.

parameters object

DashScope generation options.

parameters.duration integer

Output video duration in seconds.

parameters.resolution string

Output video resolution, for example 720P.

parameters.ratio string

Output video aspect ratio, for example 16:9.

parameters.prompt_extend boolean

Whether to enable prompt rewriting.

parameters.watermark boolean

Whether to add a watermark.

parameters.audio_setting string

Audio handling mode for video edit requests. Valid values include auto and origin.

parameters.seed integer

Random seed.

Notes

• The cURL example on this page is the Vtrix customer request format, not the raw Alibaba DashScope request format.
• Put model parameters under input[0].params.
• This model uses an async task flow. Create a task first, then poll the result by id.
• This model maps to upstream model wan2.7-i2v-2026-04-25.