Midjourney has no official, public list of banned words. It enforces a PG-13 content standard: no nudity or sexual imagery, no gore, no hateful or abusive content, no abusive manipulation of real people (including sexualized deepfakes), and nothing meant to deceive or influence elections. Prompts are checked automatically for those categories, so a single innocent word can get flagged when it often appears in blocked requests.

Disclosure: VdoBloom is our platform.

Searches for "Midjourney banned words" usually come from someone whose prompt was blocked for no obvious reason. The long word lists that circulate online are unofficial, copied between sites, and often years out of date. Midjourney judges prompts against content categories, and knowing those categories is more useful than a list. This guide covers what Midjourney's Community Guidelines prohibit (as of September 2026), why ordinary prompts get caught, what counts as a violation and what happens when you get one, and where to go if your project simply needs a different content standard.

What Midjourney's Community Guidelines actually say

Midjourney's Community Guidelines are short. The key points:

PG-13 and family friendly. Midjourney says it tries to keep the service PG-13 and admits that the AI does not always behave as expected.
Be respectful. No images or prompts that are inherently disrespectful, aggressive or abusive, and no content that is racist, homophobic, disturbing or derogatory toward a community.
No adult content or gore. Nudity, sexual organs, fixation on naked bodies, sexualized imagery and fetishes are out. So are images of people in showers or on toilets.
No gore. The guidelines define it broadly: detached body parts of humans or animals, cannibalism, blood, violence such as shooting or bombing someone, mutilated bodies, severed limbs, pestilence.
Real people. When using the Editor or uploading images, you must not manipulate images of public or private individuals in abusive, derogatory or inflammatory ways, and sexualized deepfakes are explicitly named.
No deception or illegal use. No misinformation, no images for political campaigns meant to influence elections, nothing to defraud anyone, and nothing illegal.
The spirit of the rules. Midjourney says it may act on content that breaks the spirit of the guidelines even if it is not explicitly listed.

These rules apply to everything you make, including images generated in private (Stealth) mode. Private mode hides images from the public gallery; it does not change what is allowed.

The categories that get prompts blocked

Rather than memorizing words, it helps to think in categories. A prompt tends to be blocked when its wording strongly suggests one of these:

Category	What it covers	Why Midjourney blocks it
Sexual and adult	Nudity, explicit acts, fetish content, underwear-focused or bathing scenes	PG-13 standard; shared public community
Gore and body horror	Blood, wounds, dismemberment, corpses, disease	Shock content is out of scope for PG-13
Hate and harassment	Slurs, dehumanizing depictions of groups, targeting individuals	Respect rules; harm to real communities
Real people	Abusive, sexual or inflammatory images of named or uploaded people	Deepfake and harassment risk
Deception and politics	Fake news images, election material, fraud	Misinformation and legal risk
Illegal activity	Content depicting or facilitating crimes	Legal liability
Minors	Anything sexualizing or endangering children	Zero tolerance across the industry

Midjourney has moved away from relying only on a simple word list toward automated moderation that looks at the prompt as a whole. That is why the same word can pass in one prompt and fail in another: context changes how the system reads it.

Why innocent prompts get flagged

If you were not trying to make anything on that table, the block is probably one of these false positives.

Words with a medical or anatomical meaning

Terms from biology, medicine and anatomy overlap with adult vocabulary. A prompt for an anatomy diagram, a surgery illustration or a pregnancy portrait can be read as adult or gore content.

Horror, fantasy and war art

Zombies, battle scenes, vampires and monster designs are standard concept-art subjects, but descriptions of wounds, blood and body parts land in the gore category. Midjourney's definition of gore is broader than many artists expect.

Fashion and swimwear

Swimwear, lingerie and fitted clothing are legal fashion subjects, but prompts that focus on the body instead of the garment drift toward the adult category.

Words with a double meaning

Ordinary words for animals, food, tools or weather sometimes double as slang. When the rest of the prompt is ambiguous, the moderator may read the slang meaning.

Real names

Naming a politician, celebrity or private person alongside anything negative, violent or suggestive raises the chance of a block, because that is the pattern of harassment and deepfakes.

How to rewrite a flagged prompt honestly

This is not about finding alternative spellings or hiding what you mean. Disguising prohibited content is itself treated as a violation. It is about removing ambiguity from prompts that are within the rules:

Name the genre and purpose. "Children's-book style illustration of a friendly dragon", "medical textbook diagram", "fashion catalog photo" all tell the moderator what kind of image you want.
Describe mood instead of injury. "Aftermath of a battle, smoke, broken banners, somber light" conveys war without describing wounds.
Describe the garment, not the body. Fabric, cut, color and setting work better than adjectives about skin or shape.
Drop real names from anything edgy. Use a fictional character description instead.
Change one word at a time. If a long prompt fails, remove words until it passes, so you learn which term was ambiguous.

If the image you want is truly outside PG-13, for example adult fashion with more skin or dark horror art, rewording will not and should not get it through. That is a sign Midjourney is the wrong tool for that project.

What counts as a violation, and what happens

A blocked prompt is not the same as a violation. Most blocks simply stop the job. Problems start when someone keeps pushing prohibited content or the content gets generated and reported. Midjourney's documentation says moderators can delete content that breaks the guidelines and issue warnings or timeouts, and its Terms of Service allow it to suspend or ban access, including to the Editor and Video tools, for misuse. Paid subscriptions do not protect an account from being banned. If you believe a block was a mistake, reword the prompt with clearer context, and use Midjourney's official support channels rather than repeatedly resubmitting the same flagged prompt.

When a different image model fits better

Midjourney's PG-13 line is a product choice for a public, mixed-age community. If your legitimate work sits just past it, such as swimwear catalogs, pin-up style illustration or darker fantasy, a different model is the straightforward answer. On VdoBloom's Text to Image tab each model carries a content badge. HOT models have fewer restrictions on adult, clothed content; STRICT models apply their provider's full filters. Prices per image at default settings:

Seedream 5.0 Pro (HOT): 10 credits Standard, 19 High. Closest to Midjourney's polished look. See the Seedream 5.0 Pro model page.
Seedream 4.5 (HOT): 7 credits.
Qwen Image 3.0 (HOT): 5 credits, strong with text inside images.
Qwen Image 2.1 (HOT): 3 credits at 1K, 5 at 2K.
Z-Image Turbo (HOT): 2 credits, useful for quick drafts.

Worked example: exploring a concept with twenty drafts on Z-Image Turbo costs 40 credits, then five finals on Seedream 5.0 Pro Standard add 50, for 90 credits in total. New accounts get 10 free credits. If you are rebuilding a Midjourney look, Image to Prompt turns a reference image into a written prompt you can edit. For a broader comparison, see the best Midjourney alternatives.

"Fewer restrictions" still has limits. VdoBloom checks every prompt before charging and blocks nudity and undressing requests, and any sexualized content involving minors is always blocked. If a provider rejects an image after the job starts, the credits are refunded.

Consent and real people

Midjourney names sexualized deepfakes explicitly, and the same principles apply anywhere you make images:

Use reference photos only of yourself or adults who agreed. VdoBloom requires you to confirm everyone pictured is 18+ and that you have consent before uploads are processed.
Never create images of minors in any suggestive context. VdoBloom blocks it.
Never create intimate or sexual images of a real person, public or private.
Do not produce realistic images of public figures designed to mislead.

Frequently asked questions

Is there an official list of Midjourney banned words?

No. Midjourney does not publish a complete official list, and the lists online are unofficial guesses. Its moderation evaluates the whole prompt against content categories in the Community Guidelines, so context matters more than any single word.

Why is Midjourney so censored?

Midjourney aims for a PG-13, family-friendly service shared by a large public community, so its line sits lower than many adult creative tools. Its definition of gore and adult content is broad, which catches some legitimate art.

Can you get banned from Midjourney for a blocked prompt?

A single blocked prompt usually just stops the job. Repeatedly trying to generate prohibited content, or content reported by others, can lead to warnings, timeouts or a ban under Midjourney's terms.

Does Stealth mode allow NSFW images in Midjourney?

No. Stealth or private mode only hides images from the public gallery. The Community Guidelines still apply to everything you generate.

What is a less censored alternative to Midjourney?

Models such as Seedream 5.0 Pro, Seedream 4.5 and Qwen Image are badged HOT on VdoBloom because they accept adult, clothed content Midjourney often blocks. They still refuse nudity and anything involving minors.

Midjourney Banned Words: What Gets Blocked and Why