Ideogram 4 JSON Lab

Direct structured JSON caption input for Ideogram 4. No remote magic prompt, no local Qwen prompt upsampling.

Text encoder: huihui-ai/Huihui-Qwen3-VL-8B-Instruct-abliterated

JSON caption

{
  "high_level_description":"A clean poster announcing a small experimental image generation lab.",
  "style_description":{
    "aesthetics":"minimal, precise, modern graphic design, generous whitespace",
    "lighting":"even soft studio lighting",
    "medium":"graphic_design",
    "art_style":"flat vector poster, crisp sans-serif typography",
    "color_palette":[
      "#F9FAFB",
      "#111827",
      "#2563EB",
      "#F97316"
    ]
  },
  "compositional_deconstruction":{
    "background":"A warm off-white poster background with a subtle paper texture.",
    "elements":[
      {
        "type":"text",
        "bbox":[
          250,
          130,
          430,
          870
        ],
        "text":"IDEOGRAM 4",
        "desc":"Large bold black uppercase title text centered near the upper half.",
        "color_palette":[
          "#111827"
        ]
      },
      {
        "type":"text",
        "bbox":[
          470,
          240,
          580,
          760
        ],
        "text":"JSON LAB",
        "desc":"Medium blue uppercase subtitle text centered under the main title.",
        "color_palette":[
          "#2563EB"
        ]
      },
      {
        "type":"obj",
        "bbox":[
          660,
          330,
          760,
          670
        ],
        "desc":"A thin orange rounded rectangle outline used as a design accent.",
        "color_palette":[
          "#F97316"
        ]
      }
    ]
  }
}

Mode

Turbo · 12 steps Default · 20 steps Quality · 48 steps

Width

512 2048

Height

512 2048

Main guidance

0 9

Final guidance

0 9

Seed

Randomize seed

Output

Parsed JSON caption

Compact caption string sent to model

Examples