Hybrid prompt builder — type a subject, pick from style / lighting / color dropdowns, 🎲 per field, or one-click 15 presets. Live preview shows body + --ar/--v/--s/--c flags.
PRESETS · Quick presets
SubjectDescribe the focal subject in natural English (e.g., 'a samurai in bamboo forest').
Style00Pick 1–2
Mood00Pick 1–2
Camera
Lighting
Colors00Pick 1–2
Aspect ratio
Model version
Stylize0–1000. Higher = more Midjourney aesthetic.
--s1000–1000
Chaos0–100. Higher = more variation across the 4 outputs.
--c00–100
HOW TO USE
Type the focal subject in the Subject field (English, the more specific the better)
Pick style / mood / lighting / colors — or hit '🎲 Randomize all' for a quick baseline
Set aspect ratio (4:5 for IG, 16:9 for landscapes) and version (niji 6 for anime)
Watch the prompt body's character count — Midjourney weights front-loaded words
Copy, then open midjourney.com or Discord and paste after /imagine
FAQ
Q01Is this generator free? Do I need an account?+
Completely free, runs in your browser (no API calls, nothing stored). No signup — open and use.
Q02Why are the chip labels Chinese-readable but the copied output is English?+
Midjourney does not parse Chinese prompts. The chip labels are localised for reading, but the copied IDs are always English (synthwave / golden hour / wide shot).
Q03What goes in the Subject field?+
The focal subject of your image — describe in plain English, the more specific the better. Example: 'a samurai walking through a bamboo forest at dusk'. The other fields (Style, Lighting, Colors) get appended as comma-separated modifiers.
Q04What do Stylize and Chaos do?+
Stylize (--s) controls how strongly Midjourney's own aesthetic intervenes. 0 = follow your prompt strictly. 1000 = max MJ flavour. Chaos (--c) controls how different the 4 grid outputs are. 0 = very similar. 100 = wildly different. Landscapes 250–500, product photography 0–100.
Q05Are aspect ratio (--ar) and version (--v) required?+
Aspect ratio is. Without --ar Midjourney defaults to 1:1. Version is too — without --v it uses v 7 (current latest). We force-display both so they don't get forgotten.
Q06What's the difference between v 7 and niji 6?+
v 7 is the general/photorealistic model; niji 6 is the anime model — specialised for 2D, cartoon, chibi styles. Pick niji 6 for anime; v 7 otherwise. Raw mode (v 7 --style raw) is for less-stylized output, ideal for product photography or precise realism.
Q07Why ≤ 300 chars recommended?+
Midjourney weights the front of the prompt the most; later words fade. Within 300 chars almost everything gets read. Beyond that, trailing modifiers may be ignored. The preview shows a budget bar — amber at 240+, red at 300+.