Gracias por comentar 🙌 acá tenés el recurso.
Sumale IA a tus videos.
El paso a paso para cambiar elementos de una escena que ya grabaste: tu actuación queda igual, y la IA suma lo nuevo.


El ejemplo
La misma escena. Otro resultado.
En este ejemplo cambiamos la camioneta y sumamos un personaje con IA. Tu video real es la base.


Estas capturas muestran el cambio. Para hacer el efecto cargás el video original completo, no solo una captura.
Las referencias
Elegí qué vas a cambiar.


Tip: elegí fotos donde se distingan bien los detalles del objeto y la identidad del personaje.
El paso a paso
Grabá. Cargá. Dale instrucciones.

Grabá tu videoDejá el celular quieto en lo posible. Hablá y hacé tus gestos: esa actuación es la que querés conservar.
Cargá tus archivosEntrá a la herramienta de edición de video con IA que uses y cargá el video original junto con las referencias.
Pegá el prompt y generáCopiá el prompt de abajo. Indica qué cambiar y qué conservar.El prompt, listo para copiar.
Edit @video1 using @imagen1 as the replacement vehicle reference and @imagen2 as the adult female character reference. VEHICLE: Replace the background vehicle in @video1 with the exact Rolls-Royce SUV shown in @imagen1. Use the reference image to maintain consistent visible vehicle details. Match the original vehicle's position, orientation, perspective and scale within the scene. Preserve realistic ground contact, shadows, reflections and occlusion throughout the original camera movement. The replacement vehicle is present from the first frame. WOMAN: Add ONE adult woman matching @imagen2 exactly: same facial identity, hairstyle, skin tone, body proportions and existing outfit. Use the supplied image as the identity and appearance reference for this one woman. As soon as the man starts speaking, she enters the frame from screen-left, walking naturally toward screen-right in side profile. She approaches the man and stops beside him without blocking his face or gestures. She then turns to face the camera, makes eye contact, smiles naturally and gives one small, friendly wave. After lowering her hand, she gently leans close to his shoulder in an affectionate pose for a few seconds. Adapt her position to his existing performance without moving or altering him. Keep her contact anatomically believable, with no intersecting bodies or clothing. She remains naturally alive: subtle breathing, occasional blinking and small relaxed movements. Then she straightens, turns toward screen-right and walks naturally out of frame to the right. Her exit must happen before the original video ends. She does not speak. Fit this complete action sequence within the existing duration of @video1. Use natural walking speed and smooth transitions between actions. No teleporting, sliding, sudden appearances or disappearing before she reaches the frame edge. PRESERVE THE MAN: Preserve the original man exactly as filmed: identity, face, hair, glasses if present, clothing, body proportions, hands, facial expressions, gestures, lip movements and timing. He continues his original performance without reacting differently, turning toward her or adding any new actions. PRESERVE THE SOURCE: Keep the exact original camera movement, framing, duration, aspect ratio, pacing and surroundings from @video1. Preserve the complete original audio unchanged, including his exact words, voice, synchronization and ambient sound. No additional dialogue, music, cuts, zooms or camera movements. Match the original lighting, image texture, focus, motion blur and casual smartphone-video appearance. Maintain correct depth and occlusion between the woman, the man and the vehicle. Only replace the background vehicle and add the woman performing the described entrance, greeting, brief affectionate lean and exit. No other changes. No duplicated people, identity drift, distorted faces, extra limbs, flickering or warped vehicle geometry.
Hacelo a tu manera.
La camioneta y la modelo son el ejemplo. Podés aplicar esta lógica a tus redes, a tu emprendimiento o a una presentación de producto.
Reemplazá un elemento de la escena por el producto o diseño que querés mostrar.
Sumá un personaje con una acción clara: entrar, acercarse, saludar o mostrar algo.
Partí de una toma sencilla y agregá un detalle visual que acompañe lo que estás contando.
Adaptá el prompt: cambiá el vehículo, el personaje y sus acciones. Conservá las indicaciones para mantener tu actuación, tu voz y el video original.
Antes de publicar: revisá rostro, manos, audio y que los elementos se integren bien a la escena.
¿Querés ir más allá?
Aprendé a crear con IA en la Escuela.
Un curso por suscripción para crear imágenes, modelos IA y videos, paso a paso.
— Laucha · @laucharamos10