Novo: GPT Image 2.5 Flare e Sunburst. Teste grátis com 3 imagens no cadastro; depois, 8 créditos cada.
Testar GPT Image 2.5
Espadachim de Fantasia Sombria em Dojo Chuvoso
Design de personagens

Espadachim de Fantasia Sombria em Dojo Chuvoso

Gera um retrato cinematográfico noturno de uma espadachim com rosto desfocado no pátio de um dojo chuvoso. Modifique o arquétipo, o texto do estandarte e o clima.

Prompt

Vertical cinematic dark-fantasy portrait of [a lone female martial swordswoman] standing full-body in the center of an old East Asian dojo entrance at night. She wears a torn, layered black and deep violet warrior robe with a high slit, ragged gauze-like sleeves, leather arm wraps, a wide sash belt, and simple sandals. Her long black hair is styled half-up in a messy topknot with a small ornament, with loose wet strands blowing around her shoulders. Her face is intentionally obscured by a soft rectangular blur, while the rest of the image remains sharp and meticulously detailed. She holds a sheathed katana angled downward in her right hand, hilt and tsuba guard visible near her waist. Setting: [a rain-soaked wooden dojo courtyard] with open heavy wooden double doors behind her, warm lantern light glowing from the interior, dark timber beams, drifting mist, embers, and puddles reflecting on wet stone paving. Props include: one sheathed katana held by the woman, one tattered vertical fabric banner on the right bearing the text "[武道無双]", and seven fallen bamboo practice staffs scattered across the stone floor. Lighting: [moody low-key lighting with warm lantern backlight and cool blue-gray rain ambience]. Style: photorealistic dark fantasy, detailed wet fabric textures, cinematic contrast, shallow depth of field, atmospheric, no watermark, no extra text beyond the banner.

O que você pode personalizar

Antes de gerar, troque estas partes do prompt pelo seu próprio tema ou texto.

  • [a lone female martial swordswoman]espaço reservado
  • [a rain-soaked wooden dojo courtyard]espaço reservado
  • [武道無双]espaço reservado

Detalhes do prompt

Modelo
GPT IMAGE 2
Custo por imagem
1K: 43 créditos · 2K: 86 créditos · 4K: 143 créditos
Proporção
9:16
Adicionado em

Outros modelos