High Quality AI Video from a Single Prompt
Use this 4-step workflow to generate cinematic, studio-quality AI videos — starting with nothing but a text prompt. Powered by Google Flow, NanoBanana2, and Veo 3.1.
Open Google Flow, select NanoBanana2, set image type to Editorial Portrait, and use this prompt:
High-impact cinematic depiction of Rose, featuring a defined jawline and natural complexion, short dark bob hairstyle, and intense neutral expression, dressed in a sleek black high-cut ribbed bodysuit. Positioned in a minimalist white studio, leaning back against a large glossy black exercise ball with arms raised behind her head in an athletic pose. The scene conveys confidence and poise through soft diffused studio lighting, with gentle highlights and subtle shadows. Captured using a 50mm lens, eye-level angle, shallow depth of field, cinematic framing. Ultra-detailed textures, realistic skin tones, professional cinematic color grading, sharp focus, HDR, ultra-high resolution, 4K, masterpiece quality.
cinematic, ultra-detailed, 4K, dramatic lighting, shallow depth of field, HDR, professional, film still, editorial fashion, minimalist studio, fitness aesthetic
low quality, blurry, flat lighting, distorted face, bad anatomy, extra limbs, oversharpening, text, watermark, logo, messy hair, cluttered background
Add the Step 1 image as reference in NanoBanana2 and enhance skin realism:
realistic skin texture and natural imperfections / same frame / subtle under-eye lines and shadows / uneven skin tone with slight redness and natural variation / faint visible veins / detailed fabric weave, soft wrinkles, and worn texture
Add the Step 2 image as reference and generate a high-detail face shot:
high quality close up into face / natural detailed skin texture / same camera angle / same face angle
Add both images from Steps 2 and 3 as references. Switch model to veo3.1 fast and use this prompt — replace the dialogue with your own text:
steady camera zoom in to face, model has some natural macro movements / pose remain same / model start saying this: "[REPLACE DIALOGUE HERE]"