1 Scene prompt2 Characters3 Location4 Scene settings5 Generate
1 · SCENE
0 / 2000
2 · CHARACTER REFERENCES
Add up to 3 characters. Face + full-body photos are sent as reference images so the AI keeps each person consistent. Lock continuity saves a character on this device so every future scene reuses the exact same references.
✓I confirm I have permission to use the photos, videos and voices I upload — including any real people shown or heard in them.
3 · LOCATION
AI generate Upload photo Upload video Library
The AI builds the location from your description below — name it and describe it in the notes.
Stored with the location for continuity — today’s engine uses photos + your description.
Lock location continuity — reuse this exact location in every scene
4 · SCENE SETTINGS
Aspect
16:99:161:1
Length
10s30s1 min2 min3 min
10s = one AI shot. Longer scenes are filmed as a sequence of 8-second AI shots with the same characters & location (30s ≈ 4 shots · 1 min ≈ 8 · 2 min ≈ 15 · 3 min ≈ 23), then played and edited in order. Each shot renders separately, so long scenes take a while.
Camera
AutoStaticPush-inOver-shoulderDollyHandheldDrone
Premium Streaming Quality
Auto-finish every scene — cleaned, graded & formatted for streaming
Best · AutoKling 3.0 · best facesVeo 3 · most cinematicSeedance · face matchWan 2.7 · multi-character + voiceOmniHuman · speaks your voice
Best · Auto picks the strongest engine for each scene: Kling for one character’s locked face, Wan for multiple characters or a voice clip, Veo for pure scenery.
Face lock
Keep faces true to the photos — animates the actual face instead of guessing
Strongest with one main character (their face photo is animated directly, so it stays that person). With several characters it locks each face as closely as the engine allows. For face work, the Seedance engine follows photos more closely than Veo.
AI AUTO FINISH
Finishing touches the AI applies to the generated scene.
✓AI color grade
✓Skin tone AI
✓Lighting match
✓Film grain
✓Anamorphic lens
✓Cinema bloom
✓2.39 scope framing
✓Shallow depth of field
✓AI audio · ambient sound
5 · GENERATE
Est. 2–5 minutes per 8-second shot · longer scenes film several shots
Preparing…
Locked characters & locations are saved on this device and pre-loaded next visit — so scene after scene stays consistent. AI generation runs on the deployed site.