// generator
Midjourney Prompt Generator
Enter a subject or add an image. Adjust the style and parameters, then copy your finished prompt. No sign-up.
Enter a subject or description and adjust the parameters below.
Turn an image into a prompt
Add an image, create an English description and refine it in the generator.
The image is analyzed on your device and is not sent to a server for analysis. The first use downloads the AI model (about 360–620 MB). Later uses reuse the cache if your browser retains it.
Requires a browser with WebGPU and sufficient memory. The result is a suggested description, not the original prompt. Florence-2 · MIT
Optional - add 2-3 words that sharpen the mood.
Finished prompt
Paste the finished prompt after the /imagine command in Discord, or straight into the prompt field on midjourney.com.
How to use this prompt
Four steps from finished text to a finished image.
1. Copy
Click "Copy prompt" - the text (parameters included) goes to your clipboard.
2. Paste and generate
In Discord after the /imagine command, or directly into the prompt field on midjourney.com.
3. Pick a variant
Midjourney returns four variants at once. Pick the one closest to what you had in mind.
4. Not quite right? Change one word
No need to start over - come back to the generator, change the style or lighting, and generate again.
What these parameters do
Short version - full definitions and more parameters live in the glossary.
--ar Aspect ratio
The shape of the frame. 16:9 for scenes and landscapes, 9:16 for Stories/Reels, 1:1 is universal.
--stylize Stylize strength
Lower = closer to a literal description. Higher = more artistic freedom for the model.
--chaos Variety
Higher values make the four variants differ more from each other. Good when you're searching for a direction, not a final result.
--no Exclusion
Removes something you don't want from the generation - most often text or watermarks.
Questions
Are my images sent to a server?
No. Images are analyzed locally in your browser using Florence-2. Only model files are downloaded from the internet. Descriptions may contain errors — review them before use.
Does it work on every device?
The text generator works as before. Image analysis requires WebGPU, sufficient memory and an initial model download. If analysis does not work, try an up-to-date Chrome or Edge on a computer.
Is the generator free?
Yes. No sign-up, no usage limit, no account needed.
Do I need a Midjourney account?
Not to build the prompt itself. An account (Discord or midjourney.com) is only needed once you want to actually generate the image.
Why is the prompt in English?
Image descriptions and the style, lighting and composition labels are in English. Your own subject and extra keywords stay in the language you type — the generator does not translate them automatically.
Does this replace learning to write prompts?
No. It's a starting point, not a substitute for understanding why a given prompt works - that's what the blog and 1:1 consultations are for.