wan2.6-i2v is the standard image-to-video model in the Alibaba Wan 2.6 series. It uses an image as the starting frame, then uses text to describe actions, camera movement, and plot. It supports audio-video synchronization and multi-shot storytelling, making it suitable for turning product key visuals, character illustrations, or scene concept art into short films. Compared with creation methods that only add slight motion to an image, it is better suited to visual expression with a clear beginning, development, and conclusion.
Submit requests to the public API at api.acedata.cloud using the documented parameters, then use the results in your application.
api.acedata.cloudwan2.6-i2vInput parameters and result formats vary by service. Use the public API for this model and follow its guide for generation, task retrieval and editing operations.
Duration, resolution, and frame rate are native model specifications; the platform submits requests and retrieves results through image URLs, prompts, and task APIs.
The image provides the opening composition, while the prompt explains what happens next. You can design shots around product displays, character actions, or environmental changes without having to rebuild the entire opening from text. Clearly specify the subject, direction of movement, and camera changes to make static assets the visual starting point of a short film.
Multi-shot storytelling is suitable for writing a short film as a concise shooting plan—for example, showing the environment first, then presenting the subject's actions, and finally closing on a product shot. The creative focus is not merely stacking style terms, but explaining what happens in each segment and their sequence, so limited duration can carry a clear information structure.
wan2.6-i2v has native audio-video synchronization capabilities, allowing visual actions and sound atmosphere to be considered together during creation. Prompts can describe desired scene sounds and emotions to help convey the creative goal of a sound-enabled short film. The final video should still be reviewed by watching and listening, especially to ensure that the audio matches the actions and overall rhythm.
Use a product key visual or campaign image as input, describe the camera moving closer, subject presentation, and background changes, and generate a short video for product introductions or social sharing. When creating it, first determine whether to emphasize the appearance or the usage scenario, then design actions around one key point to avoid requesting too many detail changes at once.
Use a character illustration or concept art as the starting frame, together with a brief plot description, to generate a previsualization video of action and camera progression. It is suitable for discussing character entrances, emotional turns, and scene atmosphere. The deliverable can serve as storyboard communication material, after which the creative team can choose directions worth continuing to produce.
Starting from an existing poster or scene concept image, try different actions, camera movements, and sound atmospheres to create comparable video drafts. It is suitable for discussing opening appeal and narrative order before formal production; text titles, brand marks, and calls to action can be typeset consistently in post-production for easier control of the final presentation.
If you already have a clear starting image and want to develop it into a dynamic scene, choose wan2.6-i2v. If you only have a text concept, choose wan2.6-t2v; if you want to extract a person's appearance from an existing video and use it in a new scene, consider wan2.6-r2v. The three are designed for creation based on image, text, and video references respectively, and should not be mixed simply by replacing asset fields.
wan2.6-i2v is the standard image-to-video model, while wan2.6-i2v-flash emphasizes fast generation and is suitable for first exploring multiple camera directions. Compared with wan2.5-i2v-preview, 2.6 explicitly supports multi-shot storytelling and also offers a wider native duration range. If the task requires plot progression within a short video rather than displaying a single action, 2.6 is a better fit.
Prepare a publicly accessible first-frame image, submit it using image_url, and describe the subject's actions and camera movement in prompt.
Specify model=wan2.6-i2v and action=image2video for /wan/videos, starting with 5 seconds and 720P; Wan 2.6 does not use Wan 3's 30-second duration, automatic duration, or all-purpose media parameter.
Use async=true to save task_id, query /wan/tasks or receive a callback; after completion, retrieve video_url, check the subject, actions, and sound, then proceed to editing.
Use the product image as the first frame, keep the cup centered in the frame, slowly push the camera in, slightly vary the window-side lighting, keep the background quiet, and do not switch scenes.
Confirm that the image is used as the starting point, the cup remains consistent, and the camera movement is completed; choose 720P or 1080P, and do not use Wan 3's smart duration or unified media workflow.
Use the image URL image_url to provide the starting frame, and use prompt to describe the subject's actions, scene changes, and camera movement. When calling, select model=wan2.6-i2v and explicitly set action=image2video to avoid using the default text-to-video operation. Prompts should be written as brief shooting instructions.
It natively supports videos from 2–15 seconds, with resolutions of 720P or 1080P, output as 30 fps MP4 files. When choosing, first consider whether the action can be fully expressed within the target duration; for longer stories, split them into multiple short clips and edit them together instead of cramming all plot points into a single generation.
This model natively supports multi-shot storytelling. In the prompt, describe the opening, main action, and ending frame in sequence, clearly specifying the subject, action, and transition relationship for each shot. Each segment should revolve around the same subject and plot, reducing conflicting composition requirements; the number of shots and transition effects are not locked in item by item, so check whether the transitions are natural after generation.
wan2.6-i2v natively supports audio-video synchronization and can be used to create short films with sound. You can describe the sound atmosphere and its relationship to on-screen actions in the prompt, but this is not equivalent to precise dubbing or voice cloning. Before publishing, you should still preview the finished video and check the relationship between the audio content, rhythm, and visuals.
You can set async=true to submit asynchronously and obtain a task_id, then query the final task status through /wan/tasks. Successful results provide a video link, and the response also defines video dimensions and thumbnail fields. The business workflow should first confirm that the task succeeded before proceeding to preview, save, and publish.
Choose a development approach or application suited to this model and follow its setup guide.
↗Python SDK Integration Guide
SDK installation, authentication and examples↗TypeScript SDK Integration Guide
SDK installation, authentication and examples
↗Pricing
$1.28–$4.42· per generation
You pay$13,584for 142,800 credits · $0.0951 / credits46% OFF
You're billed in credits. Pick a top-up tier to see its per-credit rate — the larger the top-up, the cheaper each credit.
This is only a basic call example. See the full docs for more parameters and advanced usage.
View Docs