Nano Banana structured JSON prompt Schema

nano-banana-structured-json-prompt

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "title": "Nano Banana (Gemini 3 Pro) Ultimate Image Schema",
  "description": "The definitive structured prompting schema for high-fidelity image generation. Includes advanced photography, multi-subject control, and text rendering.",
  "type": "object",
  "required": ["meta", "subject", "scene"],
  "properties": {
    "user_intent": {
      "type": "string",
      "description": "A natural language summary of your goal (e.g., 'Me high-fiving Batman in a neon city'). Useful for logging."
    },
    "meta": {
      "type": "object",
      "description": "Global settings for the image generator.",
      "properties": {
        "aspect_ratio": {
          "type": "string",
          "enum": ["16:9", "9:16", "1:1", "4:3", "3:4", "21:9", "3:2", "2:3", "5:4", "4:5"],
          "default": "16:9",
          "description": "The dimensions of the output.\n• 16:9: Cinematic/Desktop\n• 9:16: TikTok/Phone Wallpaper\n• 1:1: Instagram Square\n• 21:9: Ultra-widescreen Movie"
        },
        "quality": {
          "type": "string",
          "enum": ["ultra_photorealistic", "standard", "raw", "anime_v6", "3d_render_octane", "oil_painting", "sketch", "pixel_art", "vector_illustration"],
          "default": "ultra_photorealistic",
          "description": "The rendering engine or 'mode'. Use 'raw' for uncompressed photo style."
        },
        "safety_filter": {
          "type": "string",
          "enum": ["block_none", "block_few", "block_some", "block_most"],
          "default": "block_some"
        },
        "seed": {
          "type": "integer",
          "description": "Fixed number (e.g. 42) to reproduce the EXACT same image again."
        },
        "steps": {
          "type": "integer",
          "default": 40,
          "minimum": 10,
          "maximum": 100,
          "description": "Denoising steps. Higher (50+) = more detail but slower. Lower (20) = faster but rougher."
        },
        "guidance_scale": {
          "type": "number",
          "default": 7.5,
          "minimum": 1.0,
          "maximum": 20.0,
          "description": "How strictly the AI follows your prompt.\n• 1-4: Creative/Random\n• 7-10: Balanced (Default)\n• 15+: Strict/Rigid"
        }
      }
    },
    "subject": {
      "type": "array",
      "description": "List of characters or objects. Add multiple items to create multi-character scenes.",
      "items": {
        "type": "object",
        "required": ["type", "description"],
        "properties": {
          "id": {
            "type": "string",
            "description": "Unique label (e.g., 'hero', 'villain') to reference this character internally."
          },
          "type": {
            "type": "string",
            "enum": ["person", "animal", "cyborg", "monster", "statue", "robot", "vehicle", "object"],
            "default": "person"
          },
          "input_image": {
            "type": "object",
            "description": "Attach a reference photo SPECIFIC to this character (e.g., your face).",
            "properties": {
              "path": {
                "type": "string",
                "description": "Path to local file (./face.jpg) or URL."
              },
              "usage_type": {
                "type": "string",
                "enum": ["face_id", "pose_copy", "clothing_transfer", "depth_map", "full_character_reference", "style_transfer"],
                "description": "• face_id: Swaps the face (Identity).\n• pose_copy: Copies the skeleton/stance.\n• clothing_transfer: Steals the outfit.\n• depth_map: Copies the 3D shape."
              },
              "strength": {
                "type": "number",
                "default": 0.85,
                "minimum": 0.1,
                "maximum": 1.0,
                "description": "1.0 = Exact copy. 0.5 = Loose inspiration."
              }
            }
          },
          "description": {
            "type": "string",
            "description": "Visual traits: 'tall, muscular, scar on left eye, glowing cybernetics'."
          },
          "name": {
            "type": "string",
            "description": "If famous: 'Abraham Lincoln', 'Pikachu'. Triggers internal knowledge base."
          },
          "age": {
            "type": "string",
            "description": "Approximate age or life stage (e.g., '25 years old', 'elderly', 'toddler', 'middle-aged')."
          },
          "gender": {
            "type": "string",
            "enum": ["male", "female", "non-binary", "androgynous"],
            "description": "Gender presentation of the subject."
          },
          "hair": {
            "type": "object",
            "description": "Hair style and color specifications.",
            "properties": {
              "style": {
                "type": "string",
                "enum": [
                  "bald", "buzz_cut", "crew_cut", "ivy_league", "caesar_cut", "french_crop",
                  "textured_crop", "fade_low", "fade_mid", "fade_high", "skin_fade", "taper_fade",
                  "undercut", "disconnected_undercut", "side_part", "hard_part", "comb_over",
                  "pompadour", "quiff", "slicked_back", "spiky", "faux_hawk", "mohawk", "mullet",
                  "modern_mullet", "shag_cut", "wolf_cut", "butterfly_cut", "hime_cut",
                  "pixie_cut", "bixie", "bob_cut", "french_bob", "blunt_bob", "lob_long_bob",
                  "shoulder_length", "layered_cut", "long_straight", "long_wavy", "long_curly",
                  "beach_waves", "hollywood_waves", "blowout", "curtain_bangs", "bottleneck_bangs",
                  "wispy_bangs", "blunt_bangs", "afro", "short_afro", "large_afro", "dreadlocks",
                  "locs", "faux_locs", "braids", "box_braids", "cornrows", "micro_braids",
                  "twists", "two_strand_twists", "bantu_knots", "ponytail", "high_ponytail",
                  "low_ponytail", "pigtails", "bun", "messy_bun", "top_knot", "man_bun",
                  "space_buns", "half_up_half_down", "wet_look", "windswept"
                ],
                "description": "The specific cut or arrangement of the hair."
              },
              "color": {
                "type": "string",
                "enum": [
                  "jet_black", "soft_black", "dark_brown", "chestnut_brown", "light_brown",
                  "dark_blonde", "golden_blonde", "platinum_blonde", "strawberry_blonde",
                  "auburn", "dark_red", "ginger", "copper",
                  "grey", "silver", "white", "salt_and_pepper",
                  "pastel_pink", "hot_pink", "neon_green", "electric_blue", "navy_blue",
                  "royal_purple", "lavender", "teal", "rainbow", "ombre", "highlighted"
                ],
                "description": "Natural or dyed hair color."
              }
            }
          },
          "position": {
            "type": "string",
            "enum": ["center", "left", "right", "far_left", "far_right", "background", "foreground", "floating_above", "sitting_on_ground"],
            "description": "Where is this specific subject placed?"
          },
          "pose": {
            "type": "string",
            "description": "Action: 'running towards camera', 'sitting cross-legged', 'aiming weapon', 'looking over shoulder'."
          },
          "expression": {
            "type": "string",
            "enum": ["neutral", "smiling", "laughing", "angry", "screaming", "crying", "seductive", "stoic", "surprised", "tired", "suspicious", "pain"],
            "default": "neutral"
          },
          "clothing": {
            "type": "array",
            "description": "Garments worn by the subject. Separated from accessories to ensure accurate material rendering.",
            "items": {
              "type": "object",
              "properties": {
                "item": {
                  "type": "string",
                  "description": "e.g., 'bomber jacket', 'pleated skirt', 'tactical vest', 'kimono'."
                },
                "color": {
                  "type": "string",
                  "description": "e.g., 'neon pink', 'matte black', 'iridescent'."
                },
                "fabric": {
                  "type": "string",
                  "enum": ["cotton", "wool", "silk", "linen", "denim", "leather", "latex", "velvet", "chiffon", "satin", "nylon", "spandex", "lace", "tweed", "corduroy", "fur", "fleece", "mesh", "cashmere", "flannel", "sequins"],
                  "description": "The specific textile material."
                },
                "pattern": {
                  "type": "string",
                  "enum": ["solid", "striped", "plaid", "polka_dot", "floral", "camouflage", "tie_dye", "geometric", "animal_print", "paisley", "checked", "gradient"],
                  "description": "Visual pattern on the fabric."
                },
                "fit": {
                  "type": "string",
                  "enum": ["tight", "slim", "regular", "loose", "oversized", "baggy", "fitted", "flowy"]
                },
                "layer": {
                  "type": "string",
                  "enum": ["underwear", "inner", "middle", "outer", "coat"]
                }
              }
            }
          },
          "accessories": {
            "type": "array",
            "description": "Jewelry, bags, eyewear, and other non-clothing items.",
            "items": {
              "type": "object",
              "properties": {
                "item": {
                  "type": "string",
                  "description": "e.g., 'sunglasses', 'gold chain', 'tactical backpack', 'crown'."
                },
                "material": {
                  "type": "string",
                  "enum": [
                    "gold", "silver", "plastic", "wood", "glass", "metal", "diamond", "pearl",
                    "leather", "bone", "obsidian", "chrome", "fabric", "canvas", "straw",
                    "feathers", "paper", "resin", "rubber", "silicone", "ceramic", "crystal",
                    "stone", "rope", "enamel", "carbon_fiber"
                  ],
                  "description": "Material of the accessory."
                },
                "color": {
                  "type": "string",
                  "description": "e.g. 'rose gold', 'transparent'."
                },
                "location": {
                  "type": "string",
                  "enum": ["head", "face", "ears", "neck", "wrists", "fingers", "waist", "back", "held_in_hand", "floating_nearby"]
                }
              }
            }
          }
        }
      }
    },
    "scene": {
      "type": "object",
      "description": "Environment and Atmosphere.",
      "properties": {
        "location": {
          "type": "string",
          "examples": ["tokyo street", "bedroom", "mars colony", "forest", "white void", "studio", "coffee shop", "cyberpunk alley", "medieval castle"],
          "description": "The setting."
        },
        "time": {
          "type": "string",
          "enum": ["golden_hour", "blue_hour", "high_noon", "midnight", "sunrise", "sunset", "twilight", "pitch_black"]
        },
        "weather": {
          "type": "string",
          "enum": ["clear_skies", "overcast", "rainy", "stormy", "snowing", "foggy", "hazy", "sandstorm", "acid_rain"]
        },
        "lighting": {
          "type": "object",
          "properties": {
            "type": {
              "type": "string",
              "enum": ["natural_sunlight", "studio_softbox", "hard_flash", "neon_lights", "candlelight", "cinematic", "bioluminescent", "firelight", "god_rays", "rembrandt"],
              "description": "The source and quality of light."
            },
            "direction": {
              "type": "string",
              "enum": ["front_lit", "back_lit", "side_lit", "top_down", "rim_light", "silhouette", "under_lit"],
              "description": "Direction relative to the subject."
            }
          }
        },
        "background_elements": {
          "type": "array",
          "items": { "type": "string" },
          "description": "Specific items to populate the background (e.g., 'flying cars', 'cherry blossoms')."
        }
      }
    },
    "technical": {
      "type": "object",
      "description": "Virtual Photography / Camera Settings.",
      "properties": {
        "camera_model": {
          "type": "string",
          "enum": ["iPhone 15 Pro", "Sony A7R IV", "Leica M6", "Canon EOS R5", "Hasselblad X2D", "Polaroid Now", "GoPro Hero 12", "CCTV Security Cam"],
          "description": "Simulates the sensor and color science of specific cameras."
        },
        "lens": {
          "type": "string",
          "enum": ["16mm", "24mm", "35mm", "50mm", "85mm", "105mm", "200mm", "400mm", "macro_100mm", "fisheye_8mm"],
          "description": "Zoom Level.\n• 16mm: Ultra Wide/Action\n• 35-50mm: Natural/Street\n• 85mm: Portrait (Flattering)\n• 200mm: Compressed/Spy"
        },
        "aperture": {
          "type": "string",
          "enum": ["f/1.2", "f/1.4", "f/1.8", "f/2.8", "f/4.0", "f/5.6", "f/8.0", "f/11", "f/16"],
          "description": "Depth of Field.\n• f/1.2 - f/1.8: Creamy Bokeh (Blurry Background)\n• f/2.8 - f/4: Standard Portrait\n• f/8 - f/16: Everything sharp (Landscape)"
        },
        "shutter_speed": {
          "type": "string",
          "enum": ["1/8000", "1/4000", "1/1000", "1/500", "1/250", "1/125", "1/60", "1/30", "1/15", "1s", "long_exposure_bulb"],
          "description": "Motion Control.\n• 1/8000: Frozen water droplets\n• 1/60: Natural motion blur\n• long_exposure: Light trails"
        },
        "iso": {
          "type": "string",
          "enum": ["100", "200", "400", "800", "1600", "3200", "6400", "12800"],
          "description": "Grain/Noise Level.\n• 100: Clean/Sharp\n• 3200+: Gritty/Vintage/Noisy"
        },
        "film_stock": {
          "type": "string",
          "enum": ["Kodak Portra 400", "Kodak Gold 200", "Kodak Ektar 100", "Fujifilm Pro 400H", "Fujifilm Velvia 50", "CineStill 800T", "Ilford HP5 Plus (B&W)", "Kodak Tri-X 400 (B&W)", "Polaroid 600", "Kodachrome 64"],
          "description": "Color Grading Preset.\n• Portra: Warm, skin tones\n• CineStill: Halos around lights, cinematic\n• Ilford: Classic Black & White"
        }
      }
    },
    "composition": {
      "type": "object",
      "description": "Framing and Angle.",
      "properties": {
        "framing": {
          "type": "string",
          "enum": ["extreme_close_up", "close_up", "medium_shot", "cowboy_shot", "full_body", "wide_shot", "extreme_wide_shot", "macro_detail"],
          "description": "How much of the subject is visible?"
        },
        "angle": {
          "type": "string",
          "enum": ["eye_level", "low_angle", "high_angle", "dutch_angle", "bird_eye_view", "worm_eye_view", "overhead", "pov", "drone_view"],
          "description": "Camera height and tilt."
        },
        "focus_point": {
          "type": "string",
          "enum": ["face", "eyes", "hands", "background", "foreground_object", "whole_scene"],
          "description": "What is strictly in focus?"
        }
      }
    },
    "text_rendering": {
      "type": "object",
      "description": "Generate legible text inside the image (Signs, Shirts, Posters).",
      "properties": {
        "enabled": { "type": "boolean", "default": false },
        "text_content": {
          "type": "string",
          "description": "The EXACT string to write. Keep it under 5 words for best results."
        },
        "placement": {
          "type": "string",
          "enum": ["floating_in_air", "neon_sign_on_wall", "printed_on_tshirt", "graffiti_on_wall", "smart_phone_screen", "computer_monitor", "book_cover", "movie_poster", "subtitles"],
          "description": "Where should the text appear?"
        },
        "font_style": {
          "type": "string",
          "enum": ["bold_sans_serif", "elegant_serif", "handwritten", "cyberpunk_digital", "graffiti_tag", "gothic", "retro_pixel", "neon_tube"],
          "description": "The typography style."
        },
        "color": {
          "type": "string",
          "description": "Color of the text (e.g., 'glowing cyan')."
        }
      }
    },
    "style_modifiers": {
      "type": "object",
      "description": "Artistic styles and aesthetic overrides.",
      "properties": {
        "medium": {
          "type": "string",
          "enum": ["photography", "3d_render", "oil_painting", "watercolor", "pencil_sketch", "ink_drawing", "anime", "concept_art", "digital_illustration", "claymation", "papercraft"]
        },
        "aesthetic": {
          "type": "array",
          "items": {
            "type": "string",
            "enum": ["cyberpunk", "steampunk", "vaporwave", "synthwave", "noir", "minimalist", "maximalist", "gothic", "baroque", "retro_80s", "vintage_50s", "futuristic", "post_apocalyptic", "ethereal", "dreamcore", "weirdcore"]
          }
        },
        "artist_reference": {
          "type": "array",
          "items": { "type": "string" },
          "description": "Names of artists to mimic (e.g., 'Greg Rutkowski', 'Alphonse Mucha'). Use with caution."
        }
      }
    },
    "advanced": {
      "type": "object",
      "description": "Negative prompts and fine-tuning.",
      "properties": {
        "negative_prompt": {
          "type": "array",
          "items": { "type": "string" },
          "default": ["blur", "low quality", "distortion", "watermark", "text", "bad hands", "extra fingers", "mutated", "cropped", "worst quality"],
          "description": "Elements to forcefully EXCLUDE."
        },
        "magic_prompt_enhancer": {
          "type": "boolean",
          "default": true,
          "description": "If true, the AI expands your prompt with adjectives to make it prettier."
        },
        "hdr_mode": {
          "type": "boolean",
          "default": true,
          "description": "High Dynamic Range. Balances shadows and highlights."
        }
      }
    }
  }
}

https://genaiunplugged.substack.com/p/structured-outputs-json-prompts-guide

You are a data extraction specialist. Extract information from the text below and return ONLY a valid JSON object.

Schema (required fields with types):
{
  "fieldName1": "string",
  "fieldName2": "number",
  "fieldName3": "boolean",
  "fieldName4": ["array", "of", "strings"]
}

Perfect example output:
{
  "fieldName1": "example text",
  "fieldName2": 42,
  "fieldName3": true,
  "fieldName4": ["item1", "item2"]
}

Strict formatting rules:
- Output ONLY the JSON object, no preamble or explanation
- Match field names exactly (case-sensitive)
- Use empty string "" for missing text fields
- Use 0 for missing numeric fields
- Use false for missing boolean fields
- Use empty array [] for missing array fields
- Do NOT add any fields not in the schema
- Do NOT include markdown code fences or backticks
- Do NOT include the word "json" before the output
- Ensure all strings are in double quotes
- Ensure proper comma placement between fields

Validation check: Before returning, verify that all required fields are present and all data types match the schema.

Text to process:
[input text here]

You are a product data extractor. Extract product information and return ONLY valid JSON.

Schema:
{
  "productName": "string",
  "price": "number",
  "category": "string",
  "inStock": "boolean",
  "features": ["array of strings"]
}

Example output:
{
  "productName": "Wireless Bluetooth Speaker",
  "price": 79.99,
  "category": "Electronics",
  "inStock": true,
  "features": ["Waterproof", "20-hour battery", "360-degree sound"]
}

Rules:
- price must be a number without $ symbol
- category must be one word title case
- features array must contain 3-5 items
- Use false for inStock if not mentioned
- Output ONLY the JSON, no other text

Validation check: Before returning, confirm all five fields are present with correct data types.

Product description:
[paste description here]

The previous output had this error: [specific error]
Please fix it and return ONLY the corrected JSON object.
Remember: [state the specific rule that was broken]

Refusal rule:
If any required field cannot be determined from the text with high confidence, return this exact error object instead:
{
  "error": "insufficient_data",
  "message": "Cannot extract required fields from input text",
  "missing_fields": ["list", "of", "fields", "that", "are", "missing"]
}

You are an email parser. Extract key information and return valid JSON.

Schema:
{
  "senderName": "string",
  "senderEmail": "string",
  "subject": "string",
  "category": "string",
  "priority": "string",
  "summary": "string"
}

Example:
{
  "senderName": "John Smith",
  "senderEmail": "john@email.com",
  "subject": "Refund request for order #12345",
  "category": "Billing",
  "priority": "High",
  "summary": "Customer requests refund for damaged item in order 12345"
}

Rules:
- category must be one of: Billing, Technical, Shipping, General
- priority must be: Low, Medium, High, Urgent
- summary must be under 100 characters
- Use "General" for category if unclear; "Medium" for priority if unclear
- Output ONLY JSON

Refusal rule:
If email does not contain enough information, return:
{"error": "insufficient_data", "message": "Cannot classify email with confidence"}

Email to process:
[paste email here]

You are a meeting assistant. Extract action items and create a table.

Table format:
| Action | Owner | Due Date | Priority |
|--------|-------|----------|----------|

Rules:
- Due Date format: YYYY-MM-DD
- Priority must be: High, Medium, or Low
- Use "Unassigned" if no owner mentioned
- Use "TBD" if no due date mentioned
- Sort by priority (High first)
- Maximum 10 rows
- No text before or after the table

Meeting transcript:
[paste transcript here]

You are a product data extractor. Return ONLY valid JSON.

Schema:
{
  "productName": "string",
  "price": "number",
  "category": "string",
  "inStock": "boolean",
  "features": ["array of strings"]
}

Example:
{
  "productName": "Wireless Bluetooth Speaker",
  "price": 79.99,
  "category": "Electronics",
  "inStock": true,
  "features": ["Waterproof", "20-hour battery", "360-degree sound"]
}

Rules:
- price must be a number without $ symbol
- features array must contain 3-5 items
- Use false for inStock if not mentioned
- Output ONLY the JSON

Product description:
[paste description here]

https://www.productcompass.pm/p/context-engineering

https://www.apatero.com/blog/wan-2-2-training-fine-tuning-comfyui-complete-guide-2025

loras/
├── characters/
│   ├── sarah_v1.safetensors (initial training)
│   ├── sarah_v2.safetensors (retrained with more data)
│   └── sarah_v3.safetensors (current production version)
├── styles/
│   ├── corporate_professional_v1.safetensors
│   └── corporate_professional_v2.safetensors
└── motion/
    └── smooth_pans_v1.safetensors