Wan API features explained
Wan API feature overview
Wan API is an independent video API platform that gives users access to Wan and Seedance video models through both a Playground and API integration. It is set up for people who want to generate video from a prompt, an image, or reference media, then compare model options and current prices before sending a request. The platform is independently operated at wanapi.app and is not affiliated with or endorsed by Alibaba, Wan, or ByteDance.
Video models and output options
The core of the service is its model lineup. Wan API provides Wan 3.0, Wan 3.0 Prime, Wan 3.0 Pro, and Wan 3.0 Prime Pro, and each model is positioned with its own inputs and resolution choices. Wan 3.0 and Wan 3.0 Prime support output up to 1080p. Wan 3.0 Pro adds 2K and 4K output options, and Wan 3.0 Prime Pro also includes 1080p, 2K, and 4K choices.
That matters because video projects do not all need the same final quality. A short internal test clip may only need 1080p, while a finished deliverable may call for 2K or 4K. Wan API presents those options at the model level so a request can be matched to the resolution the project needs.
Ways to create video
Wan 3.0 supports text-to-video, image-to-video, and reference-to-video workflows. The homepage also describes text, image, and multimodal reference inputs, with generated sound included in the output. In practical terms, that means the same platform can start from a written prompt, a starting image, or a set of reference materials when building a scene.
The text-to-video path is useful when the idea begins as a script or scene description. The image-to-video path fits cases where a still frame already exists and movement needs to be added around it. Reference-to-video brings images, video, and audio references into one brief, which is helpful when subject, motion, and sound all need to stay aligned.
Clip length and sound
Wan 3.0 and Wan 3.0 Prime support clips up to 30 seconds, and the homepage notes generated audio for these models. The platform examples also show that sound is part of the generation workflow rather than an afterthought.
For a reader, that means the output is not limited to silent motion. A short clip with sound can be useful when a scene depends on timing, atmosphere, or spoken or musical cues, while the 30-second limit keeps the use case centered on short-form generation rather than long editing timelines.
Playground, API access, and task handling
Wan API offers both a Playground and an API integration path. The docs explain that users sign in with Google, add funds to their account, create an API key in settings, and send it in an Authorization: Bearer YOURAPIKEY header. The integration guide covers the API base URL, model ID, request fields, task status handling, result retrieval, and error responses.
This split is important because it supports two common ways of working. The Playground gives a direct place to test a model or compare requests, while the API path is aimed at sending generation jobs from software. The documentation also explains how to submit a task, poll for its status, and download the result once the video is ready.
Pricing and cost basis
Wan API shows current pricing on its model pages and pricing page. For Wan models, pricing is based on resolution and the combined duration of reference video and generated output. The pricing page also notes that text and image inputs add no video seconds, and that disabling generated audio does not lower the per-second rate.
That billing model matters because the cost is tied to video time rather than just the existence of a request. A short generation with no reference video is priced differently from a job that includes reference footage, since reference video seconds and output seconds both count toward the billable total.
What the platform is set up to help with
Taken together, Wan API is built around model comparison, short video generation, and request-based integration. A user can review the available Wan models, choose among text, image, or reference inputs, pick a resolution, and send a job through the Playground or API. The documentation and pricing pages give the pieces needed to choose a model and understand how a generation request is priced before submission.