The Core Prompt Formula
A useful Wan 3.0 prompt reads like a compact shot brief: subject and identity, action, environment, camera and framing, lighting, visual style, timing and sound.
Put the most important information first. If one detail must not change, state it directly rather than hiding it among style words.
- Subject: who or what is in the shot
- Action: the exact movement or change
- Environment: location, weather and atmosphere
- Camera: framing, lens feel and movement
- Look: lighting, color and visual style
- Audio: ambience, effects, dialogue or music
Describe the Subject Clearly
Use concrete identity details that can persist across frames. For a person, describe age range, hair, wardrobe and one or two recognizable traits. For a product, state shape, material, color and logo placement without overloading the prompt.
When reference images are present, explain which identity or design elements Wan 3.0 should preserve.
Write Action as a Sequence
Avoid vague phrases such as dynamic movement. Name the action and its direction: walks toward camera, turns left, raises a hand, then stops.
For longer clips, use chronological beats. Keep each beat visually compatible with the previous one so the model does not need to invent an abrupt cut.
Direct the Camera
Name one primary shot and one primary movement. Examples include close-up with a slow push-in, wide aerial tracking shot, handheld follow shot or static medium shot.
Too many simultaneous camera commands can conflict. If the shot changes, describe when and how the transition happens.
Prompt Native Audio
Describe the soundscape as specifically as the image. Mention room tone, weather, footsteps, engines, speech tone, music style and relative loudness.
Use negative direction when needed, such as no music or no dialogue, so ambience and effects remain the focus.
Prompt Examples
Cinematic portrait: A close portrait of a middle-aged sailor in a dark wool coat standing on a foggy pier at dawn. He turns toward the horizon as the camera slowly pushes in. Cool natural light, shallow depth of field, realistic skin texture. Distant gulls, water and soft wind, no music.
Product reveal: A matte black wireless speaker rises from a reflective surface as narrow white rim lights sweep across its texture. Slow controlled orbit, premium commercial lighting, precise reflections. Subtle mechanical sound and a restrained low-frequency pulse.
Prevent Common Failures
Remove instructions that contradict each other. Keep the requested camera motion physically possible and avoid demanding multiple unrelated locations in one continuous shot.
When text must appear in the scene, use large simple lettering and plan to finish exact typography in post-production.