HeadsUpAI

Midjourney Restores Negative Prompting to V8.1 for Precise Image Control

Midjourney, an AI image generation platform known for artistic quality, restored the --no flag for its latest V8.1 model. This anti-prompting feature allows users to specify elements they want removed, such as --no people. While a staple in previous versions, the parameter was missing from the initial Midjourney V8.1 release.

Negative prompting is essential for professional workflows where positive instructions alone cannot reliably suppress unwanted artifacts. By reintroducing this control, Midjourney is closing the feature gap between legacy models and the new architecture. This follows recent efforts to stabilize the model, including a Midjourney V8.1 image quality patch that fixed pixelation issues.

You can use the --no flag immediately across the Discord interface and official website by appending it to any V8.1 prompt. This is particularly useful for cleaning up complex scenes or ensuring style references do not introduce unintended objects. The feature is available to all subscribers as part of the standard V8.1 toolkit.

Midjourney
Midjourney
@midjourney
X

Small update today. Many requested a "anti-prompting" feature for V8 models (which existed in previous versions) which we call the --no flag. This is now available today in V8.1! So if you're trying to get something out of your images (like people) try --no people. Have fun!

8retweets141likes
View on X

Still wondering? A few quick answers below.

The --no flag is a command used in Midjourney to perform anti-prompting, which allows users to explicitly exclude specific objects or attributes from an image. By adding this parameter to a prompt, you can instruct the AI model to remove unwanted elements, such as people or specific colors, that might otherwise appear in the generation.

Yes, the --no flag is now officially available for use with Midjourney V8.1 models. While this feature existed in previous versions of the platform, it was initially unavailable when the V8 architecture first launched. This update restores the capability to the latest high-performance model, allowing for more granular creative control over image outputs.

To use negative prompting in Midjourney V8.1, you simply append the --no flag followed by the items you wish to exclude at the end of your text prompt. For example, if you want to generate a landscape without any human figures, you would add --no people to your instruction before submitting the generation request.

The update is available today for all Midjourney users currently utilizing the V8.1 model. It can be accessed through the standard Discord interface or the official Midjourney website. Because it is a core parameter update to the model itself, it does not require any special sign-ups beyond a standard Midjourney subscription.

Share this update