Wan Hosting

Wan Hosting Service is to deploying and running Wan-AI’s cutting-edge multimodal models—including Wan2.1-T2V (text-to-video), I2V (image-to-video), and VACE (video auto-captioning and editing)—on your own GPU servers. These models are available in both 1.3B and 14B parameter variants, with support for standard PyTorch and Hugging Face Diffusers formats. By self-hosting, you gain full control over generation speed, resolution (e.g., 480p, 720p), prompt privacy, and integration with your custom pipelines.

✔ GPU with at least 32-160GB VRAM (e.g., RTX 5090 / A100)
✔ Freely, Flexible Deployment Options
✔ Optional UI integrations (Gradio, FastAPI, ComfyUI)