OnlyFans restricted words are real, but there is no official list of them. OnlyFans' Content Moderation Policy says it uses “stop words”, which it describes as words banned on OnlyFans, and that text matching a stop word is automatically blocked from appearing. None of the official pages we read for this guide says which words those are. What OnlyFans does publish is the set of topics its Acceptable Use Policy prohibits, so the reliable way to write a caption, message, bio or pay-per-view description is to check it against those categories rather than against a word list someone else compiled.
This guide covers wording only. If a post or its media was taken down, the reasons are mapped in our guide to why an OnlyFans post was rejected or removed. If the whole account has been limited, go to the OnlyFans account restricted appeal checklist, which handles restrictions and appeals.
What OnlyFans officially says about banned words
Three official sources mention word-level blocking, and it helps to keep them apart because they say different things.
- The Content Moderation Policy confirms the mechanism. It lists “hashlists” and “stop words” among the technology OnlyFans uses to scan media, audio and text, and says content matching a stop word or a banned URL list is blocked automatically. A second set of scanning tools flags possible violations for human moderators to review before any decision to deactivate content.
- The interface text shows the error. OnlyFans' published English interface text contains the messages “Input contains restricted words” and a longer version that ends with a slot for the words themselves. A matching message exists for livestream comments. That is where the phrase “restricted words” comes from.
- The Acceptable Use Policy lists topics, not words. It sets out the conduct and content that may not be featured or facilitated on OnlyFans. It contains no vocabulary list.
So the accurate summary is narrower than most articles on this subject suggest. OnlyFans documents that automatic word blocking happens. It does not document which words are on the list, whether the list differs between captions and messages, or how often it changes. We read the Acceptable Use Policy, the Terms of Use, the Content Moderation Policy and the help articles cited below on 13 October 2026, and none of them publishes the stop words.
Where third-party word lists come from
Search for a restricted words list and you will find long lists on agency blogs, software sites and paid downloads. These are not OnlyFans documents. The better ones say so. One widely read example, a Supercreator guide to restricted words, states that OnlyFans doesn't publish a list and describes its own list as built with AI assistance from seed terms the company supplied, and as not exhaustive.
That tells you how to read any such list. It is a record of what its authors believe has been blocked or flagged, mixed with words they judge to be risky. A word on a third-party list may pass on your account, and a word missing from it may be blocked. Three practical problems follow.
- A list cannot tell you whether a word is blocked on its own or only in a certain context, because nobody outside OnlyFans can see the rule behind a block.
- A list can go stale without notice. Nothing in the pages we read commits OnlyFans to announcing a change to its stop words.
- A list trains you to think about spelling. The policy is about subject matter, and a caption can break it without containing a single word from anyone's list.
For that reason this guide does not reproduce a word list. The only word-level evidence worth trusting is what your own account shows you, which is what the log further down is for.
Where the wording rules apply
Creators often assume messages are private and therefore looser than the feed. The official sources say otherwise. The Terms of Use define Content as any material uploaded by a user, and the definition names text and metadata alongside photos and video. The Content Moderation Policy says OnlyFans can review and remove any content shared on the platform, and lists text and direct messages specifically. The Creator's Guide to Content Moderation repeats the point: everything posted can be moderated, including direct messages between you and your fans.
The interface text also shows how many text fields can be cleared by moderators. It includes separate notices for a username, a display name, the bio section, the website URL, the location field, a promotion campaign message, a post label and a streaming alert. Treat each of these as covered by the same rules as a caption:
- Post captions and post labels, including scheduled and queued posts.
- Direct messages and mass messages, including the description on locked pay-per-view media and any saved scripts or templates a team member sends for you.
- Profile fields: username, display name, bio, location and website link. These are visible before anyone subscribes, and the help article on explicit content applies a stricter standard to the bio and avatar for that reason.
- Promotion campaign text and livestream alerts and comments.
If someone else writes or sends text from your account, you remain responsible for it. OnlyFans' help centre says the account holder is legally responsible for ensuring that everything posted complies with the Terms of Service, whoever is helping run the account.
Pre-post caption check by policy category
Each row paraphrases a category from the Acceptable Use Policy, with detail from the moderation guide where it adds something. The third column is our reading of where that topic tends to turn up in a creator's writing; the policy itself applies everywhere. Read the draft against every row before it goes out.
| Policy category | What the policy says | Where it shows up in text | What to write instead |
|---|---|---|---|
| Age | Nothing may feature or facilitate anyone under 18. The moderation guide adds that pretending to be under 18 is not allowed even as role play or fiction. | Captions, PPV descriptions and role-play scripts that state or imply a young age, a school setting or a family relationship | Drop the scenario entirely. Describe yourself and the content as adult, with no age-play framing. |
| Illegal activity, real, claimed or role-played | The policy bans illegal activity whether actual, claimed or role-played, naming incest and sexual assault among its examples, plus anything promoting terrorism. | Story-style captions, custom-request replies in DMs and titles on locked messages | Decline the request and offer a theme that involves consenting adults with no illegal premise. |
| Consent and violence | The violence or harm category covers prohibited role play, lack of express consent, extreme impact or bondage, and suicide. | Captions that frame a scene as forced, unwilling or harmful, even when the shoot itself was consensual | Describe what happens in terms that make consent plain, and leave out harm language used for effect. |
| Meeting in person | In-person meetings are prohibited, and the help article Can I meet up with Fans? says creators and fans cannot use OnlyFans to discuss or arrange them. The moderation guide includes raffles and competitions. | DM replies to fans asking to meet, giveaway captions, bio lines about travel dates or availability | Say plainly that you only interact on OnlyFans, and run no prize draws that involve meeting. |
| Weapons, drugs and regulated goods | The policy prohibits weapons or controlled substances used in a threatening or harmful way, and separately prohibits selling controlled or regulated items. | Captions that mention drug use, offers in messages to sell or send physical items that are regulated | Leave substances out of the copy, and sell only your content through a Creator Interaction. |
| Hateful conduct | No attacking people on the basis of race, ethnicity, national origin, caste, sexual orientation, gender, gender identity, religion, age, disability or disease. | Slurs used as kink language in captions or requested by a fan in chat | Refuse the request and keep degrading themes away from protected characteristics. |
| Abuse, harassment and personal data | Stalking, doxxing and defaming are banned, as is sharing other people's personal data. | Call-out posts naming a fan or another creator, screenshots of chats quoted in captions | Raise the problem with OnlyFans support, and never name or identify the person in public copy. |
| Prohibited bodily fluids | The policy names urine and excrement as prohibited. | Menu items in a bio or pinned post, PPV descriptions, custom-request price lists | Remove the item from your menu; a word swap does not make the content permitted. |
| Inaccurate information | Misleading descriptions of media or account information are a breach. | PPV descriptions that promise more than the locked media contains, bios claiming something the account does not offer | State what the buyer receives: who is in it, roughly how long it is and what it shows. |
| Links and off-platform references | The cyber activity category covers spamming, linking to external media storage sites, and referencing an off-platform site that violates the Terms of Service. The moderation guide also lists shoutout-for-shoutout content linking to an SFS account. | Cloud-storage links in messages, the website field, repeated identical promo messages, shoutout captions | Deliver media inside OnlyFans, and check any link against the policy before it goes in a bio or message. |
| Transactions outside a Creator Interaction | The in-person meetings clause extends to facilitating any transaction other than a Creator Interaction, which the Terms define as a subscription, a pay-per-view payment, or another interaction or payment between a user and a creator's account or content. | DMs offering a discount for paying through another app, bio text listing outside payment handles | Price everything as a subscription, tip or locked message on the platform. |
| Endorsement and other people's rights | You may not represent that OnlyFans has endorsed you or your content, or infringe a third party's intellectual property. | Bio badges such as “official” or “OnlyFans approved”, captions built on a brand name or a celebrity's name | Describe your own ranking or niche in your own words, without implying OnlyFans or a brand backs you. |
Two labelling rules in the Terms of Use work the other way: they require words rather than ban them. Advertising content must carry a signifier such as #ad, #paidpromotion or #sponsored in the caption, with the brand name and who paid. AI-generated content must be captioned with a signifier such as #ai or #AIGenerated. Leaving these out is a wording problem too.
What to do when a message or caption is blocked
A block at the point of sending is the stop-word mechanism working as the Content Moderation Policy describes. Nothing has been posted, so there is nothing to appeal. Work through it in order.
- Read the error. The interface text has a version of the message with room for the words concerned, so the screen may tell you what triggered it. Copy the message exactly before you change anything.
- Find the category. Use the table above to decide which policy topic the wording touches. If you can't find one, the word may be on the list for a meaning you didn't intend.
- Decide whether the idea is allowed. If the underlying content or offer breaks the policy, the answer is to remove it, in the copy and in the media.
- If the idea is allowed, rewrite the sentence in plain terms that make the permitted meaning obvious, then send it once.
- Record the result in the log below, including the wording that went through.
- If you still can't tell why something was blocked, the moderation guide says to ask by emailing support@onlyfans.com. Our guide to contacting OnlyFans support as a creator explains which route handles which question.
Don't respell your way past a block. Swapping letters for symbols or numbers changes what the filter sees and leaves the meaning intact. The policy governs the meaning. The same Content Moderation Policy describes scanning tools that surface text for human moderators, and lists consequences that run from deactivating content through warnings to deactivating the account. A disguised word that a moderator later reads in context is a worse position than a blocked message.
The blocked-wording log
Because no official list exists, your own record is the best reference you will have. Keep it in a spreadsheet the whole team can read, and add a row every time text is blocked, cleared or deactivated. Over a few months it becomes a house style guide based on what happened on your account.
| Log column | What goes in it | Why you will want it later |
|---|---|---|
| Date and sender | When it happened and which team member was typing | Shows whether one person or one script keeps hitting blocks |
| Text field | Caption, DM, mass message, PPV description, bio, display name, post label or stream comment | OnlyFans doesn't say whether every field is checked the same way, so your log is how you find out |
| Blocked wording | The full sentence as first written | Context matters more than the single word |
| Error shown | The on-screen message copied exactly, including any words it named | It is the only official statement of what was matched |
| Policy category | The row of the check table it most likely falls under, or “unclear” | Separates real policy problems from accidental matches |
| Decision | Idea removed, or idea kept and reworded | Stops the same prohibited offer being redrafted next month |
| Wording that passed | The sentence that was accepted, if one was sent | Gives writers and chatters an approved phrasing to reuse |
| Follow-up | Any later notice about the same text, and any reply from support | Passing the filter is not the same as being approved |
Review the log before any bulk send. Our mass message pre-send checklist covers the rest of that routine, and the wider rulebook is summarised in the OnlyFans compliance guide.
Blocked text versus deactivated text
The two outcomes are easy to confuse. A block stops text before it appears. A deactivation removes something that was already live, after automated scanning or a moderator found a problem. The help article Why was my content deactivated? says you are sent a notification explaining the reason, and that the reasons are based on the Acceptable Use Policy.
This is why passing the filter proves little. The stop-word list is one tool among several, and the Content Moderation Policy says OnlyFans can review any content at any time. A caption that sent without an error can still be deactivated later if its meaning breaks a policy category. Deactivations can be appealed through the Deactivation Appeal Form, and the steps for that sit in the two guides linked at the top of this page.
Limitations of this guide
The limits here are significant, and they come from what OnlyFans chooses not to publish. We cannot tell you which words are on the stop-word list, and nobody outside OnlyFans can verify a claim that a particular word is banned. We have not confirmed in a live account what the restricted-words error displays; the description above rests on the published interface text. The third column of the check table is our interpretation of where each policy category meets everyday copy, and a moderator may read a sentence differently.
The policy summaries reflect the pages as they stood on 13 October 2026. The Acceptable Use Policy was marked as last updated in September 2025 and the Content Moderation Policy in May 2026, and OnlyFans can change either. The policies are the binding text, so read them directly before relying on a paraphrase. This is general information and not legal advice; where a question turns on the law in the place you live or film, ask a lawyer there.
SirenCY, the publisher of this guide, is an OnlyFans management agency based in Melbourne. Our OnlyFans management agency page describes how agency support is arranged.