顶尖模型,工具内置。
搭配 GPT、Claude、Gemini、Grok 等模型,搜索网络、深度研究、读取文件、生成图片、运行代码。
上传一张带有你想要造型的图片,免费获得一条提示词,在 ChatGPT 或 Gemini 中把这种造型用到你自己的照片上,而且不改变你的脸。
把参考图拖到这里
上传你想模仿的造型图,不是你自己的照片。
无需账户即可试用。免费账户每天可生成 2 条照片编辑提示词。 Zemith 的免费工具。Zemith 是集聊天、图片和视频于一体的 AI 应用。
好的 ChatGPT 或 Gemini 照片编辑提示词,会先说明哪些不能变:你的脸、你的特征以及照片里的每一个人,然后才描述要改什么。否则两款应用都会连同其他部分一起重画你的脸。这个工具写的每条提示词都以这种方式开头。
你不必自己描述造型,只要把它展示出来。工具会从参考图中读取场景、服装、姿势、光线、色彩和镜头感,并用文字写出来,因为 ChatGPT 和 Gemini 只会看到你的照片。它只描述造型,从不描述参考图中的人,所以结果看起来仍然是你。
想要现成的提示词? 20 条保持面部不变的 ChatGPT 照片编辑提示词
这个工具根据参考图真实生成的提示词。

提示词
Using the uploaded photo as the reference, keep the person's face, facial features, identity, skin tone and body shape exactly as they are in the uploaded photo, and keep them fully clothed and recognizable. Restyle the scene into this look: a street-style fashion portrait taken outdoors in front of a pale concrete parking structure with horizontal slatted grilles, a metal wire fence and a low concrete kerb, the ground light grey asphalt. The outfit is a crisp oversized white button-down shirt left untucked, a thin black tie loosely knotted, a short pleated camel-tan skirt, fine white fishnet tights and black knee-high block-heel platform boots. Pose the person mid-stride, stepping forward with one leg lifted behind, torso angled away and head turned back over the shoulder toward the camera with a calm confident expression. Hair styled in long tight braids falling forward over the shoulders. Bright hazy backlight with warm sun flare bleeding across the top of the frame, soft washed-out highlights, warm cream and beige color grade, light film grain and a slightly faded analog finish. Subject placed slightly off-center, full body in frame, vertical 4:5 portrait framing. Photorealistic result with natural skin texture and clean, undistorted hands and face.

提示词
Using the uploaded photo as the reference, keep the person's face, facial features, identity, skin tone and body shape exactly as they are, and keep their expression natural. Recreate the whole scene and look of the following description with the person from the photo. The person sits on the roof rack of a dark SUV, perched near the front edge with one leg bent and the other foot resting lower against the side of the car, hands relaxed, head tilted down and slightly away, calm and contemplative. Long hair blows sideways in the wind. Outfit: a loose black sleeveless top over a black long-sleeve layer, dark ripped jeans with a torn knee, and worn grey-and-white sneakers. Setting: open countryside at sunset, a field of tall dry corn stalks and grasses in the background with thin silhouetted posts on the horizon, the car windows catching warm reflections. Golden-hour backlight from low behind the subject, strong rim light along the hair and shoulders, deep shadowed silhouette of the body against a soft gradient sky of pale cream fading to warm orange. Warm film-like color grade, slight grain, gentle lens glow and flare, shallow depth of field, shot from a slightly low side angle with the car filling the lower half and the sky the upper half, wide horizontal framing around 3:2. Photorealistic, natural skin texture, sharp hands and face, no distortion.

提示词
Using the uploaded photo as the reference, turn it into a flat vector-style illustration while keeping the person's face, facial features, identity and body shape fully recognizable, and keeping everyone in the photo as they are without adding or removing anyone. Draw the person seated in profile on a simple four-legged wooden stool, facing left, shoulders slightly rounded, hands resting near the knees, one leg bent with the foot flat on the ground, the other tucked back. Dress them in a light beige short-sleeved button-up shirt with a collar and chest pocket, a dark navy pleated knee-length skirt, dark navy knee-high socks and brown lace-up shoes; hair falls in a straight chin-length bob. Place them low in the frame against a large plain warm sand-beige background. Behind them rises a tall irregular vertical rectangle of swirling stylized night sky in layered flat shapes: deep navy, cobalt, teal, pale blue and cream clouds, a round pale-yellow moon low in the middle, small yellow and cream four-point stars and scattered dot specks floating past the panel's torn edges. Use crisp clean outlines, flat unshaded color fills, minimal texture, and a soft diagonal cast shadow on the ground. Square 1:1 framing, person small in the lower center.

备注: Only the lighting and the dark background, keep my own outfit
提示词
Using the uploaded photo as the reference, keep the person exactly as they are: the same face, facial features, identity, skin tone, body shape and expression, and keep their own outfit unchanged, including its colors, cut and layers. Only change the lighting and the background. Recreate a moody low-key studio portrait look: the background becomes a plain dark charcoal-black wall, matte and slightly uneven in texture, with faint subtle scuffs and marks and no objects or props. Light the person with a single soft directional lamp from slightly above and in front on one side, so one side of the face and body falls into gentle shadow while the other is lit with warm, natural skin tones. Keep the shadows deep but open, with soft gradual falloff and no harsh edges, and let the darkness wrap around the shoulders so the figure separates cleanly from the near-black wall. The overall palette stays dark and desaturated, cool in the shadows with a faint warm cast in the highlights, low contrast and quiet, serious mood. Finish with realistic photographic detail, natural skin texture, crisp focus on the face, gentle depth of field, no distorted features or hands, vertical 4:5 portrait framing from the waist up with the person centered and calm space above the head.
添加一张带有你想要造型的图片:潮流照片、服装、姿势、光线或场景都可以。如果只想要其中一部分,写一句简短备注。
点击生成提示词。你会得到一条描述造型并保持你面部不变的提示词。免费账户每天可生成 2 条。
把提示词复制到 ChatGPT 或 Gemini,并附上你自己的照片,或在 Zemith 的图片编辑器中打开。
想在一个地方完成?点击在 Zemith 中帮我完成,会打开 Zemith Creative Studio 的图片编辑器,并已填好提示词。你在那里添加照片并选择模型,例如 Nano Banana Pro 或 GPT Image。照片编辑属于 Zemith 付费方案,每次编辑前都会显示费用。
FAQ
花不到一个 AI 订阅的钱,就能用上 GPT、Claude、Gemini 等模型,外加 50 多款 AI 工具。
受信赖的企业团队
顶尖模型,加上图像、视频和语音工具,一个套餐全包含。
没有 Zemith
分别付费的总计US$234.70/月
"I love the way multiple tools they integrated in one platform. Going in the right direction."
— simplyzubair
"The quality of data and sheer speed of responses is outstanding. I use this app every day."
— barefootmedicine
"The credit system is fair, models are perfect, and the discord is very responsive. Quite awesome."
— MarianZ
"Just works. Simple to use and great for working with documents. Money well spent."
— yerch82
"The organization of features is better than all the other sites — even better than ChatGPT."
— sumore
"It lives up to the all-in-one claim. All the necessary functions with a well-designed, easy UI."
— AlphaLeaf
"The team clearly puts their heart and soul into this platform. Really solid extra functionality."
— SlothMachine
"Updates made almost daily, feedback is incredibly fast. Just look at the changelogs — consistency."
— reu0691
把调研、写作、设计和跟进工作交给 Zemith。它会自行选择所需工具,交付完成的成果。
签这份租房合同前,我应该检查什么?
已读取租房合同.pdf · 12 页
搭配 GPT、Claude、Gemini、Grok 等模型,搜索网络、深度研究、读取文件、生成图片、运行代码。
研究我的前 3 大竞争对手并撰写报告
Zemith 会在云端继续工作,完成后通知您。
把我的发布笔记整理成任务、一张发布配图和一条跟踪表记录
已完成。事项已创建,跟踪表已更新
Notion、Linear、Canva、Airtable 等。创建或更改任何内容前,都会先征得您的同意。
把我的 Q3 笔记整理成董事会汇报
好的。我会把您的笔记整理成董事会文档、演示文稿、KPI 表格和 PDF。
文档、幻灯片、表格和 PDF,随时可发送。
在可视化画布上串联模型和工具,从一条提示词到一支完整的宣传视频。
简报、报告和提醒按计划运行,在您需要时已准备就绪。
正在共享屏幕您“能帮我检查一下作业吗?”
Zemith“我看到的答案都对,您笔下的 3 × 4 等于 12。”
实时语音,能看到您的摄像头或屏幕。



顶尖图像和视频模型,尽在一个工作室。
将 PDF、链接和 YouTube 视频变成播客、测验、抽认卡和思维导图。