Photogpt
Enhance the uploaded image into an ultra-sharp, highly photorealistic 4K photograph while preserving the original image exactly. Keep everything in the source photo unchanged: the same woman, recognizable face and identity, hairstyle, clothing, body proportions, pose, hands, trophy, people at the edges, room, display cabinet, flags, lighting, shadows, reflections, colors, framing, camera angle, perspective, and overall composition. Do not redesign or reconstruct the scene. Treat the original image as the absolute visual reference and only improve its image quality. Recover and naturally reconstruct the facial details that are obscured by the low resolution and haze while keeping the same person and facial structure. Preserve the original expression and hairstyle. Restore realistic skin texture, hair strands, fabric details, trophy reflections, glass reflections, background objects, and environmental details that are already present in the source. Remove the heavy blur, haze, compression artifacts, pixelation, noise, and softness while maintaining the original photographic character. Improve sharpness and clarity without making the image look oversharpened or artificially generated. Use realistic DSLR-quality detail, natural skin texture, physically accurate lighting, subtle depth, and authentic photographic grain. The trophy must remain exactly where it is, with the same shape, position, reflections, and appearance. Keep all background elements and people exactly as they appear. Do not change the woman’s clothing, body shape, proportions, pose, expression, or appearance. Do not add or remove anything. Final result: ultra-realistic 4K restoration of the exact original photograph, as if the same moment had originally been captured with a modern professional camera. Negative prompt: new person, changed face, face reconstruction that changes identity, changed hairstyle, changed clothing, changed pose, altered body proportions, altered trophy, removed background elements, added objects, different room, different lighting, beauty filter, plastic skin, excessive sharpening, CGI, cartoon, illustration, AI-looking face, oversaturated colors, text, watermark.

PhotoGPT 的AI 图片

GPT Image 2
提示词

Enhance the uploaded image into an ultra-sharp, highly photorealistic 4K photograph while preserving the original image exactly. Keep everything in the source photo unchanged: the same woman, recognizable face and identity, hairstyle, clothing, body proportions, pose, hands, trophy, people at the edges, room, display cabinet, flags, lighting, shadows, reflections, colors, framing, camera angle, perspective, and…

尺寸2304px x 1728px
参考图片
源图片 1

开始探索

加入拥有数千名活跃用户的 Discord 社区，了解最新提示词灵感并探索社区作品

生成图像加入 Discord
Google Veo 3.1Seedance 2