Real outputs, shown with their inputs.

A photorealistic original 10-second two-performer rap duet, one continuous locked medium-wide shot, landscape 16:9. The FIRST reference image (image 1) is ONLY the LEFT performer, the orange tabby cat with green eyes; the SECOND reference image (image 2) is ONLY the RIGHT performer, the black-and-white border collie dog with brown eyes. Preserve each subject's exact face, eye shape, hair or fur markings and clothing independently throughout. Two separate subjects remain visible at all times. Never swap their left/right positions, never blend or merge their identities, never duplicate or replace either subject. Pure saturated orange seamless photography studio background and floor (#F47721), exactly ONE hanging microphone centered between the two performers. 0-4 seconds LEFT performs original rhythmic rap mouth movements and small expressive gestures while RIGHT listens and nods; 4-8 seconds RIGHT performs while LEFT listens; 8-10 seconds both nod to the beat. Natural anatomy and small controlled movements, faces unobstructed, no crossing, no camera motion or cuts. Original choreography only, no real artist, celebrity, copyrighted performance, song, logos or text. Generate an original simple beat and original playful vocals, no recognizable song.

A photorealistic original 10-second two-performer rap duet, one continuous locked medium-wide shot, landscape 16:9. The FIRST reference image (image 1) is ONLY the LEFT performer, the dark-skinned adult man with short black curls, beard and teal shirt; the SECOND reference image (image 2) is ONLY the RIGHT performer, the auburn-haired freckled adult woman in navy shirt. Preserve each subject's exact face, eye shape, hair or fur markings and clothing independently throughout. Two separate subjects remain visible at all times. Never swap their left/right positions, never blend or merge their identities, never duplicate or replace either subject. Pure saturated orange seamless photography studio background and floor (#F47721), exactly ONE hanging microphone centered between the two performers. 0-4 seconds LEFT performs original rhythmic rap mouth movements and small expressive gestures while RIGHT listens and nods; 4-8 seconds RIGHT performs while LEFT listens; 8-10 seconds both nod to the beat. Natural anatomy and small controlled movements, faces unobstructed, no crossing, no camera motion or cuts. Original choreography only, no real artist, celebrity, copyrighted performance, song, logos or text. Generate an original simple beat and original playful vocals, no recognizable song.