Make Xianxia Videos with Veo 3—Hits Every Time (Plus Veo 3 Prompts)
Xianxia (cultivation fantasy) clips keep performing on short-video platforms. With Veo3 image-to-video, you lock face and costume from a reference still, then steer motion with clear Veo prompts—a workflow that scales. Below: four real setups and prompts you can paste.

Why xianxia content travels well
The genre ships with built-in visuals: misty peaks, glowing artifacts, flowing robes. Audiences already know the mood. Swap sects, storylines, or heroines and the same pipeline still works.
What Veo 3 image-to-video fixes
Text-only Veo3 runs often drift on faces and East-Asian styling. Image-to-video anchors the reference look—hair, robes, accessories—then your prompt handles motion. That consistency matters for character-led xianxia shorts.
Four cases and Veo 3 prompts
Use English prompts for image-to-video (better pass rates on most platforms). Table below is copy-ready.
| Case | Visual focus | Veo 3 prompt (English) |
|---|---|---|
| Huangfeng Valley · Chen Qiaoqian | Speaks to camera, wind in robes, sad turn and walk away | The woman looks directly into the camera, lips slightly parted as if speaking. Her hair and clothes sway gently in the wind. Then she turns away sadly and walks off slowly. |
| Spirit pet | Slow push-in, tearful silent sigh | The camera pushes in very slowly toward her face. She maintains eye contact with the lens, but her expression grows sadder, lips slightly parted as if sighing silently. A tear continues to slide slowly down her cheek. |
| Sect junior | Friends splashing water, lotus smile, handheld feel | The shot starts on her face with a calm expression. Women in the background playfully splash water toward her; she breaks into a bright smile. Then someone swims up beside her and waves happily. The woman in the foreground picks up a lotus flower and smiles at the camera. Camera movement should feel natural and handheld with slight shake. |
| Fairy in the city | Selfie POV, turn through busy street | The self-held camera starts on her calm face. She turns gracefully, back to camera, as if moving through a crowd. She turns again to face the lens and keeps walking through a busy street. People around her add depth and motion. Natural handheld feel with slight shake. |
How to run image-to-video on Veo 3
| Entry | Notes |
|---|---|
| Gemini web | Google account → 2.5 Pro → Video → Add photo → English prompt (Veo 3 Fast, ~8 s) |
| Google Flow | Pro credits: Fast ~20 pts/clip, Quality ~100 pts/clip |
Practical tips
| Item | Advice |
|---|---|
| Reference image | Some uploads get filtered—try another crop or angle |
| Prompt language | Prefer English for image-to-video |
| Aspect ratio | 16:9 landscape works best |
| Audio | Veo 3 can add ambience—hint wind, sigh, water splash in the prompt |
Where the edge still is
Tools save time; shot design and character bible still differentiate you. Fix a template—reference still + one motion beat + one camera move—then batch and pick winners.
Try Veo3 image-to-video for xianxia shorts directly in your browser.
Start using Veo