Image-to-Video Generation with Wan 2.2

★ 5.0 · n8n · complex · 2 nodes

Server-side image-to-video on Wan 2.2 is an n8n automation template that, together with a Python application, deploys a self-hosted video generation service from images. Reach for it when you need to animate a photo or illustration without sending data to third-party cloud services, when full control over generation infrastructure matters, or when GPU-accelerated speed is a priority. Trigger: formTrigger. Integrations: Modal, Hugging Face, Wan 2.2, TurboDiffusion. The Python server runs on Modal with an H100 GPU and uses the Wan 2.2 I2V A14B model via TurboDiffusion — 1 to 4 inference steps, 81 frames, 16 fps, 720p with adaptive aspect ratio; the auxiliary text model UMT5-XXL and weights are pulled from Hugging Face. The n8n workflow contains 2 nodes: a form where you upload an image and enter a prompt, then the workflow sends both to the server and returns a finished MP4. The archive includes the n8n workflow and the Modal server code. Modal and Hugging Face accounts are required. Direction — image-to-video generation, not text-only video synthesis. Suitable for teams ready to run their own GPU stack on Modal; not a fit if you need a no-setup SaaS solution.