Wan 3.0 Video
wan3.0-videoCommercial useWan 3.0 Video is Wan's text to video model, available here through a single relay API. It generates a short video clip from a written description. It takes a text prompt as input. Requests run server-side against 97AI.PRO with your credit balance, so you get Wan quality without managing a separate Wan account or key.
Text prompt describing the desired video. Supports Chinese and English, up to 20,000 characters.
Optional first-frame image URL for image-to-video.
Optional last-frame image URL. Use together with first_frame for first-to-last-frame video.
Reference image URLs for all-in-one reference mode. Refer to them as Image1, Image2, etc. in the prompt.
Reference video URLs for all-in-one reference mode. Refer to them as Video1, Video2, etc. in the prompt.
Reference audio URLs for voice, music, or ambience guidance.
Optional public file URL, such as a PPTX, PDF, or other supported media file.
Optional web link used as source material.
Output video resolution tier.
Output aspect ratio. adaptive lets the model choose from the input and prompt.
Output duration in seconds. Use 2-30 for generated clips, or -1 for smart duration mode.
Whether the generated video includes audio.
Random seed from 0 to 2147483647.
Whether to add a watermark.