This workflow generates 9 campaign images from your reference and model images. Here's how to set it up:

To make this easier, We’ve created a full template for the workflow, upload your reference image and your model, and it’s ready to generate. See the template below.

Go to a tool like ChatGPT, upload your reference image, and add the prompt below. It will analyze the image and generate nine different angles that match the same vibe and environment.

{"instruction":"Generate a single high-resolution image in 3:4 aspect ratio (portrait) featuring a professional 3x3 grid (9 individual frames). Analyze the provided reference image and preserve the exact same model identity, face, body, outfit, clothing details, styling, hair, makeup, accessories, pose language, and overall presence. Preserve the exact same environment, location type, background structure, spatial depth, surface materials, and atmosphere from the reference image. Preserve the exact same lighting logic including direction, intensity, softness/hardness, shadow behavior, contrast structure, and natural or artificial light feel. Preserve the exact same Photographic Rendering Profile including lens feel, focal compression, perspective behavior, depth of field, focus falloff, sharpness, micro-contrast, dynamic range, highlight roll-off, shadow depth, contrast curve, color science, color grading, temperature, saturation, tonal bias, grain/texture cleanliness, and overall rendering identity. Across all 9 frames, do not redesign the outfit, do not change the location, do not alter the lighting direction, do not modify the rendering style, and do not introduce new props or new environments. Only vary camera angle, framing, pose, expression, composition, and perspective while keeping every other visual characteristic locked to the reference image.","layout":"3x3 portrait grid in 3:4 aspect ratio with no outer margin. Frames extend fully to the edges. Thin 2–3px black (#000000) gaps only between frames. Each frame taller than wide.","shot_list":["Frame 1: Eye-level tight portrait of the exact same model from the reference image, preserving identical identity, outfit, styling, environment, and lighting. Direct gaze, calm editorial expression, centered composition, clean background separation, same rendering profile as the reference image.","Frame 2: Slight 45-degree angle mid-shot of the same model in the exact same environment. Preserve all clothing, styling, and lighting characteristics. Slight head turn, subtle expression shift, off-center composition, same lens feel and depth behavior as the reference image.","Frame 3: Full-body portrait in the exact same location showing more of the environment and spatial depth visible in the reference image. Same model, same outfit, same styling, same lighting direction. Relaxed but polished pose, balanced editorial composition, identical rendering consistency.","Frame 4: Low-angle portrait of the same model in the same setting, keeping all environmental and lighting characteristics unchanged. Slightly stronger pose, composed expression, fashion-editorial body language, same color grading and photographic finish as the reference image.","Frame 5: Straight-on hero portrait as the anchor frame of the 3x3 grid. Same model, same outfit, same styling, same background, same atmosphere, and identical lighting logic. Strongest eye contact and presence, symmetrical composition, premium editorial framing, exact same rendering identity.","Frame 6: Side-profile portrait of the same model in the same environment with no changes to wardrobe, location, or light structure. Refined posture, contemplative expression, elegant composition using background structure from the reference image, same depth of field and tonal response.","Frame 7: Slight overhead angle portrait within the same scene, using the same model and same styling exactly as shown in the reference image. Pose adjusted for visual variation, subtle expression change, use of negative space and environmental framing, identical lighting and rendering profile.","Frame 8: Dynamic close-to-mid portrait of the same model in the same environment with a slight shift in body posture or movement. Preserve exact outfit, same facial identity, same hair and styling, same lighting behavior, and same commercial/editorial rendering profile. More expressive composition while remaining cohesive.","Frame 9: Wider environmental portrait in the exact same location with the same model, same clothing, same styling, and same atmosphere. Use foreground/background depth and strong compositional framing while keeping the model as the focal point. Confident polished pose, identical lighting logic, and identical photographic rendering profile."],"style_notes":"Cohesive professional photoshoot. Identical environment, lighting logic, and Photographic Rendering Profile across all 9 frames. High-end editorial quality. Strong visual continuity. Analyze the attached reference image carefully and preserve the exact same character model with the exact same clothing, styling, environment, atmosphere, and visual rendering across every portrait frame."}

This will give you a new prompt with the full JSON instruction to create the campaign.

Now use go to Gemini and select Nano Banana 2 (Nano banana works but wont give as good results as Nano Banana 2.
Hit Generate.

AI Campaign workflow

Refrence-to-campaign