Master WAN with our step-by-step tutorial, detailed feature walkthrough, and expert tips.
Explore the key features that make WAN powerful for video generation workflows.
WAN (wan.video) is an AI video generation platform developed by Alibaba's Tongyi Qianwen (Qwen) team, the same group behind the Qwen large language model series. It is built on the open-sourced Wan 2.x family of video foundation models, which were released in 2025 and have been positioned as one of the leading open video generation models. The platform exposes more than 40 generative abilities, ranging from text-to-video and image-to-video to specialized tasks like sketch-to-video and video super-resolution. It is hosted on Alibaba Cloud infrastructure, giving it access to large-scale GPU compute.
WAN operates on a freemium model with a free tier and pay-as-you-go paid usage billed through Alibaba Cloud credits. The free tier provides a limited daily generation allowance for core tasks like text-to-video, image-to-video, and text-to-image at no cost. Paid usage is billed per generation through Alibaba Cloud's DashScope API pricing: standard-resolution text-to-video clips (typically 4â5 seconds) cost approximately $0.12â$0.20 per clip (~ÂĨ0.24 per second of generated video at 480p), while higher-resolution outputs and advanced abilities like video super-resolution cost more, roughly $0.25â$0.50 per clip at 720p+. Image-to-video and sketch-to-video are priced in a similar range. A light creator generating 5â10 clips per day might spend approximately $3â$8 per month, while a moderate production user running 20â40 generations daily could expect $15â$40 per month. This compares favorably to Runway's entry plan at ~$15/month (which includes a fixed credit bundle) and Pika's ~$10/month starter tier. However, because WAN uses variable per-generation pricing rather than a flat subscription, actual monthly costs depend directly on usage volume, resolution choices, and which abilities are used.
WAN supports a wide range of video generation modes, including text-to-video (generate from a written prompt), image-to-video (animate a still image), sketch-to-video (turn a rough drawing into motion), and speech-to-video (drive a character or scene from audio). It also offers post-generation tools such as video extension to lengthen an existing clip, video repainting to restyle a video, video composite edit, and video super-resolution to upscale output quality. This breadth makes it suitable for short-form social content, product animations, and creative experiments alike.
Compared to Runway, WAN offers a much broader menu of image and video abilities in a single interface, while Runway has a more polished editor and stronger ecosystem integrations. Versus Pika Labs, WAN is better suited for users who want one platform for both image and video work. Against OpenAI's Sora, WAN's advantage is open access today plus a free tier, whereas Sora is gated and US-centric. Compared to Kling, WAN has stronger backing from a hyperscale cloud (Alibaba Cloud) and an open-source model lineage, which is meaningful for developers and researchers.
Commercial use is generally permitted under WAN's terms when content is generated through a paid plan or an account in good standing, but rights and restrictions can vary by region and ability type. Since WAN is operated by Alibaba, the terms of service follow Alibaba Cloud's content and IP guidelines, which require that prompts and outputs do not infringe third-party rights. For high-stakes commercial campaigns, users should review the latest terms inside the console and confirm licensing for any specific ability they rely on. For the open-source Wan 2.x models themselves, license terms on the model release should be checked separately.
Now that you know how to use WAN, it's time to put this knowledge into practice.
Sign up and follow the tutorial steps
Check pros, cons, and user feedback
See how it stacks against alternatives
Follow our tutorial and master this powerful video generation tool in minutes.
Tutorial updated March 2026