
GPT Image 2.0
Minimalist hand-drawn illustration of a politician at a podium with microphones, black lines on paper
Prompt
Generate a minimalist, childlike hand-drawn image on paper around a specific subject: first extract the subject from the complete information of a full photograph-like depiction, compressing it into a few jittery black lines that remain recognizable, the minimum identifiable outline, and a small number of key symbols, then let a large area of clean empty space support it. The subject should be understood through its outer contour, posture, holes, edges, relationships between objects, and one or two handwritten marks; the lines should look as if drawn on paper with a black ink pen or thick pencil, with uneven thickness, slight buildups of black at turns, circles carrying the deviations of a trembling hand, long lines drifting slightly, and interior details reduced to only the necessary short strokes. Reduce the complexity of the objects to the threshold of recognition: remove one stroke and they become difficult to identify, add one stroke and they become too refined. Clothing, utensils, plants, buildings, food, machines, or abstract symbols should all be transformed into white filled shapes, black edges, loose connecting lines, and one small distinctive feature that best conveys their identity. If people appear, keep their faces paper-white, using only dot-like eyes, a short nose line, and minimal expression; compress hair, beards, clothing folds, hands, and feet into black filled blocks, broken lines, or loose contours, so the person looks like a remembered symbol rather than a portrait. The background is a paper field or a pale flat field, serving the functions of breathing room, scale, and quiet observation, with clear empty space preserved above and around the sides. Place a single subject in the center or slightly below center; compress multi-element content into a low horizontal band of relationships, linking it with a few trembling baseline strokes so the empty space above remains silent. Let color submit to the black lines and empty space: a white or near-white background should occupy most of the area, black lines should carry the structural authority, low-saturation pale colors should be used only to group forms and convey emotional warmth, and small areas of warm or dark color should serve only as semantic accents; the coloring should be slightly uneven, with the edges subtly misaligned from the contours, preserving exposed paper white, grain, and a hand-applied feel. If text appears, it should look like a low-fidelity label or small note written on the surface of the object with the same black pen, with uneven lettering, loose spacing, and slight misalignment, while remaining legible; the text belongs to the object itself rather than to an external layout. Overall, maintain the tension between low finish and high recognizability, like a casual drawing that was carefully left behind: sparse, light, clumsy, readable, with a slight scanned-paper feel, ink residue, broken lines, repeated tracing marks, and uneven coloring traces.
围绕具体主题内容生成一张极简稚拙的纸面手绘图像,先把主题从完整照片式信息中抽离出来,压缩成几笔仍能认出的抖动黑线、最低识别轮廓和少量关键符号,再让大片干净空白托住它。主体靠外轮廓、姿态、孔洞、边缘、物体关系和一两个手写标记被读懂;线条像黑色墨水笔或粗铅笔画在纸上,粗细不均,转折处略有堆黑,圆形带着手抖的偏差,长线微微漂移,内部细节只留下必要短线。对象的复杂度要降到识别临界点,少一笔会难认,多一笔会太精,衣物、器具、植物、建筑、食物、机器或抽象符号都转成白色块面、黑色边缘、松散连接线和一处最能说明身份的小特征。若出现人物,脸保持纸白,只用点状眼睛、短鼻线和极少表情,头发、胡子、衣褶、手脚都压成涂黑块、断线或松散轮廓,让人物像被记住的符号而不是肖像。背景是纸面场或浅色平场,承担呼吸、尺度和安静观看的功能,上方和四周保留明显空白;单主体落在中部或偏下位置,多元素内容压成低位横向关系带,用几条颤动底线串联,让上方空场保持沉默。色彩服从黑线和空白,白色或近白底占最大面积,黑线拥有结构权,低饱和淡色只用于形体分组和情绪温度,小面积暖色或深色只作为语义重音;填色略不均匀,边缘轻微错开轮廓,保留纸白漏出、颗粒和手工涂抹感。文字若出现,应像同一支黑笔写在物体表面的低精度标签或小注,字形不齐、间距松、略有错位,但仍能读懂,文字属于物体本身而不是外部版式。整体保持低完成度与高清识别之间的张力,像一张被认真留下的随手画,空、轻、笨拙、可读,带轻微扫描纸感、墨水残差、断线、重复描痕和非均匀涂色痕迹。
The rights to this prompt belong to its author @xiaoxiaodong01. You are responsible for any commercial use — promoting AI tools, aggregators or Telegram bots: mozgi.io accepts no liability. Details →











