Blonde noir · in motion
One original portrait reference. One recorded image-to-video output. Compare the source framing with the finished clip, then use the still and an editable motion prompt to start a separate request.
The starting image
The input is a portrait-format AI visual reference. The recorded output is landscape, 1280 × 720. That change in framing matters when you evaluate what a reference preserves.
Compare the face, clothing, lighting and edges across the clip. A reference guides generation; it does not lock every feature or guarantee frame-by-frame identity.

About this sample

- Original input
- One reference image
- Sample workflow
- Image to video
- Sample model
- Wan 3.0 Spicy
- Output
- 6 s · 1280 × 720 · MP4
- Playback
- 30 fps · 16:9 · Silent
These settings describe this recorded sample. Select an available model and review its options in Studio.
Original generation record
This reference was submitted through WaveSpeed using Wan 3.0 Spicy. It is an externally generated brand study, not a Studio credit-usage test.
Task trace and evidence
- Record reference
- RAP-68cc899225f28c48
- Creation route
- WaveSpeed API, outside Studio
- Recorded status
- Completed provider response recorded
- Model identifier
- alibaba/wan-3.0/image-to-video-spicy (recorded provider model identifier).
- Local record time
- Billing evidence
- A provider quote is retained in the source record. An actual billed amount and a Studio credit transaction are not verified.
Time the local polling result reported completed; exact provider completion time is not retained. A local record time is not a first-publication date or a guaranteed generation-time measurement.
Evidence checked: Provider task submission and polling records; Original request settings; Matching displayed-file hash. Task references are shortened fingerprints; private task identifiers and response URLs are not published.
This record is not linked to a Studio generation or credit transaction. It does not establish the same behavior for a currently listed model.
Original prompt and actual settings
A single continuous six-second cinematic fashion portrait of the adult blonde woman in the first frame. Preserve her facial identity, blue eyes, blonde hair, black jacket and opaque black top throughout the entire shot. All clothing stays fully in place. The camera makes an extremely slow smooth push-in. She makes one slight natural head turn and a relaxed blink, with ordinary composed nonsexual movement. Wine-red velvet curtains move very gently in the background. Soft neutral light on her face, restrained red contour lighting, detailed almost-black shadows, natural skin texture and consistent red-and-black noir film color grading. Keep comfortable headroom and shoulders within frame. Professional fully clothed editorial portrait. Silent, no cuts, no text, logos, UI or watermark.
- duration
- 6
- resolution
- 720p
- aspect ratio
- 16:9
- seed
- 20261115
- generate audio
- false
- enable prompt expansion
- false
These are the recorded request values. Current Studio options and credits can differ.
This sample is currently available in Preview and has not been publicly published. A first-publication date has not been assigned.
What you can reuse
The Studio link prepares the still reference and a suggested motion prompt. You can edit the prompt, replace the picture and select an available model. It does not copy motion from the video or automatically submit a request.
Read the suggested motion prompt
Animate this adult portrait with a subtle head turn, a natural blink and gentle breathing. Keep the black eveningwear, red curtain background and cinematic lighting consistent. Use a slow, steady camera move and avoid abrupt cuts.
This is an editable starting suggestion, not the original generation recipe. The sample describes one recorded result, not a comparative test or a promise of identical output. How we document examples.
Check the current options
The model catalog can change. The sample’s six-second output is a recorded fact, not a promise that the same duration is selectable in every current model. Compare current image-to-video configurations and plans and credits before your next request.