Updated September 2026
Wan 3.0 on FunPhantom
Wan 3.0 on FunPhantom is reference-to-video: lock identity with 1–5 images, optional short voice samples, native audio. You run it in the studio; this page explains how.
How Wan 3.0 runs here
Upload 1–5 reference images. Images below 240×240 are rejected. Optional voice samples are labeled man or woman. Jobs consume tokens. Remix can reuse a public listing without stacking hidden prompt wraps.
- Open FunPhantom and sign in.
- Go to Generate and choose Wan 3.0 / Video+Audio.
- Add 1–5 reference images (at least 240×240). Optionally attach a man or woman voice sample.
- Write the motion prompt, spend tokens, and wait for the job to finish.
Visuals are AI-generated. Do not upload real people without consent.