
<lora:BArtstyle-KREA2-V1-1:0.3>, BArtstyle style,<lora:blinklikeer_k2v1:0.5>, 4x0style, <lora:training_3876538-20260630144051811:0.5>, zy_cinematic, <lora:Krea 2 - Chiaroscuro Lighting Style:0.5> ,chiaroscuro lighting style, <lora:michiru_v1_000001000:0.8>, michiru, 1girl, solo, short black hair, blue hair tips, multicolored hair, ahoge, purple eyes, red maple leaf hair ornament, black choker, black crop top, sleeveless top, midriff, denim shorts, belt, white feather accessory, black thigh straps, black leg warmers, black platform sandals, open-toe footwear. VERTICAL 2:3 COVER COMPOSITION. Michiru is the only human character visible. She occupies approximately sixty-five percent of the image. Position her slightly right of center and leave the upper-left sky and cherry blossoms available for title placement. Michiru sits sideways on the low stone wall beside a riverside cycling path. A city bicycle is parked directly beside her without blocking her legs. She supports herself with one hand on the wall, while her other hand holds a small bottled drink. One leg hangs naturally over the wall and the other knee bends slightly. She looks at the camera with a relaxed, satisfied smile and faint blush. The riverside contains cherry trees, falling petals, grass, flowerbeds, benches, cycling signs, metal railings and a paved path. Across the river are apartment buildings, small houses, a pedestrian bridge, utility poles and a distant railway line. The calm water reflects the spring sky and buildings. Soft morning light passes through the cherry blossoms and creates dappled shadows. Keep petals away from Michiru’s eyes and distinctive accessories. Clearly show both sandals and her visible toes. Unified Japanese anime illustration, detailed background and cohesive spring colors. AVOID: additional cyclists, pedestrians, crowd, duplicate Michiru, bicycle covering character, photorealism, cropped feet, text, logo, watermark.
Parameters used to generate this content
Copies the original metadata text embedded by AUTOMATIC1111.
AI models used to generate this content