Midjourney's V8 model family launched without one of the most-used tools in the toolkit: the --no parameter. After enough user requests, it's back. The --no flag is now available in V8.1, letting you explicitly tell the model what to leave out of a generated image.
The feature is simple in concept: append --no to any prompt, followed by a comma-separated list of things you don't want.
It gives Midjourney a direct instruction to steer clear of specific concepts, and it's the most reliable way to get clean, focused results.
Think of it as creative subtraction rather than addition.
Under the hood, it's a weight
Using the --no parameter is the same as using a -0.5 weight.
So vibrant tulip fields --no red is functionally identical to vibrant tulip fields:: red::-0.5. The flag is just a cleaner shorthand for a concept that power users were already doing manually with multi-prompt syntax.
Negative weights give you more control over how strongly certain words influence the final image -- when you add one, you're telling Midjourney to push an element away rather than remove it completely.
Don't miss what's next in AI
Join 300,000+ engineers and researchers who get the signal, not the noise.
- Full access to in-depth AI research breakdowns
- Be the first to know what's trending before it hits mainstream
- Daily curated papers, repos, and industry moves