Create

Generate Similar

Generation Mode Choose how the image should start.

Prompt

{
  "generation_request": {
    "meta_data": {
      "tool": "NanoBanana Pro",
      "task_type": "photorealistic_museum_romance_editorial",
      "version": "v1.0_MUSEUM_KISS_PAINTING_PARALLEL_SCENE_IDENTITY_LOCK",
      "priority": "highest"
    },
    "references": {
      "couple_reference_image": {
        "slot": 1,
        "purpose": "EXACT_COMPOSITION_ENVIRONMENT_POSE_LOCK",
        "strict_lock": true,
        "no_layout_drift": true,
        "preserve_framing": true,
        "preserve_pose": true,
        "preserve_camera_angle": true,
        "preserve_background": true
      },
      "female_reference_image": {
        "slot": 2,
        "purpose": "FEMALE_IDENTITY_LOCK",
        "strict_identity_lock": true,
        "face_similarity_priority": "MAX",
        "no_face_morph": true,
        "no_beautify": true
      },
      "male_reference_image": {
        "slot": 3,
        "purpose": "MALE_IDENTITY_LOCK",
        "strict_identity_lock": true,
        "face_similarity_priority": "MAX",
        "no_face_morph": true,
        "no_beautify": true
      }
    },
    "output_settings": {
      "aspect_ratio": "4:5",
      "orientation": "portrait",
      "resolution": "ultra_high_res",
      "render_style": "ultra_photoreal_editorial",
      "sharpness": "crisp_but_natural",
      "film_grain": "subtle_analog",
      "color_grade": "clean_gallery_neutral"
    },
    "hard_constraints": [
      "Exactly 2 people only (one female, one male).",
      "Faces must match the uploaded female/male references with maximum similarity.",
      "Recreate the reference image composition EXACTLY: museum gallery, blue wall, framed painting behind the couple, placards below the painting, floor and lighting consistent.",
      "Pose must match exactly: the couple kissing, hands on each other’s face/neck/shoulders in the same placement, bodies aligned the same way.",
      "No additional people, no crowd in background.",
      "No readable text on placards (keep them blurred/unreadable), no logos, no watermark.",
      "Photorealistic anatomy (hands/fingers), natural skin texture, no beauty filter.",
      "Painting content stays visually similar in framing and placement; keep it non-readable and non-identifiable beyond general impression."
    ],
    "creative_prompt": {
      "scene_summary": "Ultra-photoreal romantic museum photo, recreated exactly from the reference. A couple stands in front of a deep blue gallery wall and a large gold-framed painting. They are kissing intimately, with the man’s hands gently holding the woman’s face and the woman’s hands resting naturally on his shoulder/neck area, matching the reference pose precisely. The painting behind them depicts a romantic embrace, visually echoing the couple’s pose.\n\nEnvironment:\n- Museum interior with a deep blue wall, one large gold-framed painting centered behind the couple, and small museum placards directly beneath it.\n- Warm gallery spotlights create soft, realistic shadows on the wall and subtle reflections on the frame.\n- Stone/terrazzo museum floor visible at the bottom with the same perspective and horizon line.\n\nCamera & realism:\n- Natural documentary style, like a real candid museum photo.\n- 35mm–50mm lens look, medium depth of field: couple sharp, background slightly softer but still detailed.\n- Preserve natural skin texture and realistic lighting; no stylization, no CGI."
    },
    "negative_prompt": [
      "extra people",
      "crowd",
      "face morph",
      "identity drift",
      "beautify filter",
      "cartoon",
      "anime",
      "cgi",
      "plastic skin",
      "deformed hands",
      "extra fingers",
      "blurred faces",
      "readable text",
      "watermark",
      "logo",
      "oversharpening",
      "overexposed highlights"
    ]
  }
}

3808 / 10000

Based on: Romance in a Gallery: A Photorealistic Embrace

Model

Select a model to continue.

Each model exposes different quality and editing capabilities.

Size & Aspect Ratio

Pick the output framing.

Available options change with the selected model.

Select a model to see credit cost. Reference mode adds 1 credit on supported models.

Generate Similar

Ready when you are

Generating your image

Image ready

We could not finish this image

Generating your image

Image ready

We could not finish this image

Send people to the full prompt and image