68个 Midjourney 高频 Prompt 提示词

- holographic
- in the style of kawaii aesthetic
- colorful garden
- in a style of surrealism
- light silver and light gold
- neon light
- pastel accent colors
- in the style of circular abstraction
- atmospheric lighting
- candid moments captured
- in the style of conceptual installation
- emotional and dramatic scenes
- in the style of surreal engine
- carpetpunk
- troubadour style
- celebration of rural life
- liquid foil gradient
- euphoria
- in the style of minimalist monochromatic landscapes
- surreal fashion photography
- post-internet aesthetics
- in the style of hyperrealistic photography
- in the style of organic material
- algeapunk
- transfixing marine scenes
- roman art and architecture
- in the style of pastel color palette
- in the style of unreal engine
- in the style of hyperrealistic animal portraits
- adventurecore
- neon sparks
- nature-inspired camouflage
- rural life
- unicorncore
- hyperbolic expression
- kodak
- muted tones
- ektachrome
- in the style of organic forms
- neon colors
- forestpunk
- cambodian art
- in the style of psychadelic surrealism
- in the style of dark reflections
- in the style of gauzy atmospheric landscapes
- conceptual installation art
- product photography
- dreamy
- in the style of surreal architectural landscapes
- light white and light amber
- mirror rooms
- dark pink and aquamarine
- macro photography
- pastel colours
- in the style of fractal geometry
- national geographic photo
- in the style of realistic hyper-detailed portraits
- geometric minimalistic
- the circle mirror in the mirror
- california impressionism
- in the style of neo-pop sensibility
- in the style of dreamy symbolism
- polaroidcore
- in the style of minimalist surrealism
- murals and wall drawings
- in the style of monochromatic minimalist portraits
- in the style of solarization
- foil holographic

Midjourney 是一个由 Midjourney 同名研究实验室开发的人工智能程序，可根据文本生成图像，于2022年7月12日进入公开测试阶段，目前以公司形式运营，使用者可通过 Discord 的机器人指令进行操作，输入 Prompt 提示词后可以创作出多种多样的图像作品。

Midjourney AI 与其他 AI 模型类似，它使用海量图像来利用机器学习的方式进行训练。当用户输入 Prompt 提示词时，它的机器人会找到与描述相匹配的图片，并将它们进行艺术组合以创建独特的图像。

我从 MJ Discord 频道中抓取了 15,943 条信息，将 JSON 上传到 ChanGPT 代码解释器，提取出 13,682 个 prompt 并按照受欢迎程度对它们进行了排序，并选出了其中的 68 个。

@ciguleva

以下是来自 @ciguleva 的68个高频 Prompt 提示词: