Midjourney prompt blocked: How to bypass safely

Midjourney uses an automated moderation system to prevent the creation of unsafe, offensive, or copyrighted content. However, this system frequently flags completely harmless prompts due to false positives, homonyms, or over-sensitive keyword matching.

If you receive a red warning message stating that your prompt is blocked or violates community standards, you can resolve the issue without risking an account ban by following these steps.

1. Locate and Remove Red-Flag Words Midjourney does not publish a list of banned words, but its filters are highly sensitive to specific categories. Your prompt may contain a word that has a dual meaning or is associated with a restricted topic.

1. Check your prompt for terms related to violence, bodily fluids, physical harm, weapons, or political figures.
2. Look for innocent words that contain flagged letter combinations or could be interpreted sexually (e.g., "breast" when describing a chicken breast, or "strip" when describing a strip of paper).
3. Remove any direct names of modern political figures, controversial historical events, or highly guarded intellectual properties.

2. Strip the Prompt to Isolate the Trigger If you cannot figure out which word triggered the block, you can isolate it by using a process of elimination.

1. Divide your prompt into halves.
2. Try running the first half of the prompt on its own.
3. If it generates without a block, the problem word is in the second half.
4. Subdivide the problematic section until you identify the specific word or phrase triggering the auto-moderation system.

3. Rewrite Sensitive Concepts with Literal Descriptions Instead of using direct, abstract, or dramatic nouns that might trigger filters, describe the scene using neutral, literal visual elements.

• Instead of: "A bloody battle scene with soldiers"
• Use: "A historical painting of soldiers, red paint splatters, dramatic cinematic lighting, historical armor"
• Instead of: "An explosive, chaotic car crash"
• Use: "A vehicle with dented metal frame, smoke rising, scattered glass on the ground, high-action motion blur"

Focusing on textures, colors, and lighting rather than violent verbs or nouns allows you to achieve the desired mood without triggering the safety filter.

4. Use Multi-Prompts to Separate Flagged Compound Words Sometimes, two words placed next to each other trigger the filter, whereas separating them allows them to pass. You can use multi-prompts (using double colons `::`) to split the concepts for the AI parser.

1. If a phrase like "killer whale" is flagged due to the word "killer," you can structure it as orca or try separating the weight: whale:: killer::0.1 (though using the direct alternative name is always safer).
2. Use synonyms where possible to avoid complex multi-prompt workarounds. For instance, replace "explosion" with "burst of bright light and flying debris."

5. Submit an Appeal for False Positives If your prompt was blocked by the automated system but is entirely safe and compliant with Midjourney’s Community Guidelines, you can appeal the block directly within Discord.

1. Look at the system message containing the block notification.
2. Click the Appeal button directly below the warning message (if available).
3. The prompt will be sent to a human moderator or a secondary, more advanced evaluation model. If approved, your prompt will generate within a few minutes.

When to Escalate Do not repeatedly try to bypass filters with minor spelling changes (e.g., "bloody" to "bl00dy"). Repeated violations of the Terms of Service can lead to a temporary or permanent account suspension. If your account is mistakenly locked or flagged for prompt violations, go to the official Midjourney Discord and post in the `#moderation-appeals` or `#member-support` channels to request manual review from a staff member.