Wan × MuleRouter
Wan Models are powerful multimodal models that transform text and images into high-fidelity videos and images.
Access premium, exclusive, and multimodal variants through MuleRouter's unified API.
Wan Collections
Browse the full library of Wan text-to-video and image-to-video models.
Interactive Playground
Sign in to try Wan 2.5 Text-to-Image generation directly in your browser.
Playground Setup
Output
{
"task_info": {
"id": "9e8587c5-6634-40e9-8b7c-d5c9d276ab34",
"status": "completed",
"created_at": "2025-11-24T11:45:40.765320+00:00",
"updated_at": "2025-11-24T11:46:08.050930+00:00"
},
"images": [
"https://www.mulerouter.ai/wanresult_00.webp",
"https://www.mulerouter.ai/wanresult_01.webp",
"https://www.mulerouter.ai/wanresult_02.webp",
"https://www.mulerouter.ai/wanresult_03.webp"
]
}1{
2 "model": "wan2.5-t2i-preview",
3 "prompt": "Minimalist square monochrome photography, realistic, soft natural light, a simple landscape with strong contrast and clean lines, subtle sense of wind in the scene, fine-grain texture, gallery-level, high-end editorial style.",
4 "size": "1280*1280",
5 "n": 4,
6 "prompt_extend": true
7}Wan Models on MuleRouter
Explore our complete lineup of Wan models for text-to-video, image-to-video, and multimodal generation. All models available through MuleRouter's unified APIs with enterprise-grade reliability.
Full-Spectrum AI Models
Efficient, reliable, and multi-mode AI text/vision-gen model API, support a full-range of multimodal tasks, including image generation, image editing, video generation and video editing.
High-Speed Inference
Built on AI infrastructure, our APIs deliver orders-of-magnitude faster inference through parallel processing, distributed acceleration, and dynamic resource scheduling.
Developer-Friendly Integration
Flexible integration via HTTP and SDKs. Supported, allowing seamless connectivity with just a few lines of code, accept local files as a file, base64 strings, and public URLs.
Collections built around the Wan roadmap.
Each block bundles models that share schedulers, training data, and latency expectations so your team can mix and match without surprises across the Wan core lineup.