Streamlabs lets you filter words in Alert Box messages with its profanity filter and custom bad-word list. If you need a person to judge a message before it appears, use alert moderation instead: filtering acts on matching words or alerts, while moderation holds alerts for approval.
These controls apply to Streamlabs alert text, not automatically to every message in YouTube live chat. Set up the filter, choose whether to censor or suppress a matching alert where the current dashboard offers that choice, then save and test before relying on it during a live broadcast.
What alert word filtering controls
An alert overlay can display text associated with an event, such as a viewer message or a name attached to a follow or subscription. Word filtering is meant to limit what appears in that alert. Streamlabs documents both a built-in profanity filter and a custom list of words or phrases you want to block.
A word filter is a rule, not a human judgement. It can act on terms you expect to see, but it cannot reliably understand every spelling, context, or variation. A phrase may be harmless in one message and unwelcome in another; a filter that matches it has no way to weigh that context for you. Do not treat the setting as a guarantee that no unwanted text will appear.
The relevant controls are in Streamlabs Alert Box settings. They are separate from YouTube’s live-chat moderation controls. YouTube Help describes blocked words for live chat in YouTube Studio, but that is not the same thing as filtering text rendered by a third-party alert overlay. If you manage both, configure and check both systems rather than assuming one passes its rules to the other. See YouTube’s live chat moderation guidance for the platform-side controls.
It helps to decide what you want to happen to a matching alert. Censoring a word can preserve the event while obscuring the matched text. Disabling an alert containing a bad word can hide the whole alert, including a legitimate event or other useful message. The exact wording and available choice can change, so inspect the current dashboard rather than relying on a remembered label.
Open Alert Box settings
Sign in to Streamlabs and open Streaming essentials > Alert Box. In the Alert Box settings, look for the general settings area; this is where the documented profanity and custom-word controls are configured. Streamlabs’ quick guide to setting up alerts describes the Alert Box path and the profanity filter.
If the navigation looks different from a guide or screenshot, check that you are in the Streamlabs dashboard and have opened Alert Box rather than a separate widget or an account-level notification page. Dashboard labels can evolve. The useful destination is the Alert Box’s general settings, not YouTube Studio’s live-chat moderation page.
Make a note of the current behaviour before changing it, particularly if the channel is already running. A saved change can affect what viewers see in subsequent alerts. If you have a second moderator or editor, agree who will review the settings and test them; a filter list that one person updates without telling the rest of the team can be difficult to diagnose later.
For a channel that plays a prepared programme continuously, the alert overlay is only one component of the viewing experience. The same care you give the alert should extend to the loop itself; for example, the guide to running pre-recorded nature videos as a YouTube live stream covers the content side rather than alert moderation.
Enable the profanity filter
In Alert Box general settings, enable Profanity Filter. Streamlabs describes this as a control that blocks bad words from appearing in alerts. Depending on the current interface, you may be able to use its standard list, a custom list, or both. Read the options in the dashboard before choosing, as the exact labels and scope may change.
The built-in list is useful when you want a general starting point without compiling every term yourself. It is not a promise of complete coverage, and you should not assume it recognises every spelling, language, transliteration, or deliberate alteration. If your audience uses several languages, test realistic examples for the language and writing styles that appear in your channel’s alerts.
Decide whether a match should be censored or whether the alert should be disabled, if that choice appears in the current settings. Censoring leaves the event visible but changes the offending text; disabling the alert can remove the entire item from view. That difference matters on a devotional channel, for instance, where you may want a genuine subscription alert to remain visible even if its attached message needs to be obscured.
There is no measured accuracy figure in the cited Streamlabs guidance, so do not plan around a particular success rate. Keep the decision practical: use the built-in filter for a general first pass, and add specific recurring terms that are inappropriate for your channel. For messages that require interpretation, a filter alone is the wrong tool; use approval moderation as well.
Add custom bad words
Use Custom Bad Words in the Alert Box general settings to enter the words or phrases you specifically want blocked. Add only terms that are genuinely unsuitable for your channel. A broad or ambiguous term may also appear in ordinary names or harmless messages, so think through the likely false matches before saving it.
Streamlabs’ dedicated instructions caution creators to consider whether a banned word could be part of a normal message. This is important where names, place names, transliterated words, or devotional vocabulary overlap with an unwanted term. Start with clear, recurring cases instead of building a very broad list on the assumption that more entries always mean better moderation.
A custom list works best for known patterns, not for every way a viewer might phrase something. Users can change spelling or substitute characters, while an innocent message can contain a matching sequence. Review actual alert examples and adjust the list carefully. If a term blocks an otherwise useful alert, remove or refine it and test again rather than adding more terms blindly.
Keep a brief record of why each custom entry was added and who maintains the list. This is especially useful when more than one person operates the channel. A note such as “repeated unwanted phrase in alert messages” is enough to explain why a rule exists without preserving offensive text in a shared document. Revisit entries when the audience or the channel’s content changes.
Save and test the settings
Save the settings before leaving the page. Then use the dashboard’s available preview or test controls to check the alert presentation. A test should verify the behaviour you selected, not just that an alert can appear: check whether a matched term is censored or whether the whole alert is suppressed, and confirm that an ordinary event still displays as you expect.
Do not test a newly changed rule for the first time during an important live moment. Use harmless placeholder examples that let you see where the alert appears and how the text changes. If the dashboard’s test feature does not reproduce a particular viewer-message path, treat that as a limitation of the test and keep a moderator ready to respond to a real alert.
Testing also helps catch a common configuration mistake: editing a list but navigating away without saving. Reopen the general settings after saving and confirm that the filter and custom terms are still present. If you make several changes, change one thing at a time and check the result; otherwise you may not know which setting caused an unexpected alert to disappear.
Keep the overlay readable at the size viewers actually see. A channel that is watched on a phone may need a shorter alert message or a less intrusive layout, independently of its word filtering. The article on running recorded tuition videos around the clock from a laptop in India addresses a different part of operating a long-running channel, but the same principle applies: check the viewer-facing result rather than assuming a dashboard setting tells the whole story.
Use moderation for manual review
Filtering and moderation solve different problems. A filter applies a rule to matching text; alert moderation holds alerts for a person to approve before they appear. Use the former for repeated terms you already know you want to block. Use the latter when context matters, the messages are unpredictable, or you want someone to review alerts before they reach the overlay.
Streamlabs’ setup instructions describe enabling alert moderation and holding alerts for approval, with regular or unlimited approval-delay options. Choose the delay behaviour shown in the current dashboard based on whether someone can actually review alerts while the channel is live. A queue that waits for approval is only useful if a moderator knows where to find it and is available to act.
Manual review costs attention. During a quiet overnight stream, there may be nobody at the controls; an approval queue can leave legitimate alerts waiting. Decide in advance who will monitor it and what to do when no one is available. You can use automatic filtering for clear, recurring terms and reserve manual approval for messages where a human decision is worth the interruption.
| Approach | What happens | Useful when | Main trade-off |
|---|---|---|---|
| Profanity or custom-word filter | Matching language is censored or, depending on the selected setting, the alert is disabled | The unwanted term is known and repeated | Rules can miss variations or catch harmless text |
| Alert moderation | An alert is held for approval before appearing | Context needs a person’s judgement | Someone must check and approve the queue |
The table is a distinction, not a guarantee that either method catches every unwanted message. For a 24/7 channel, consider the hours when a moderator is absent and keep expectations realistic. If you use YouTube chat moderation as well, remember that YouTube’s blocked-word list governs live chat rather than automatically configuring Streamlabs alerts. A practical overview of how to create a call-in show for a live stream is also relevant if your format depends on viewer contributions, because people and moderation procedures become part of the programme.
Troubleshoot text that still appears
First identify where the text appeared. Was it inside a Streamlabs alert overlay, in YouTube live chat, or in a different widget? A YouTube chat setting does not necessarily alter text that a separate alert system displays. Likewise, the Streamlabs Alert Box filter is not a substitute for the platform’s chat controls.
Next check that the profanity filter is enabled and the custom entry is saved in the Alert Box general settings. Confirm that the rule is intended to censor text or suppress the full alert; a censored word can still leave the rest of the message visible, which may look as if the filter did not act. If you chose suppression, check whether the entire alert was simply absent rather than partly changed.
Consider spelling and context. A filter may not match a variation, spacing, or substituted character in the way you expect, and a word can appear as part of an innocent name or phrase. Do not infer that the filter has failed from one example without checking the exact text, the selected behaviour, and whether the alert came through the control you configured.
If an ordinary alert disappears, review recently added custom terms first. Streamlabs notes that a banned word may be part of a normal message. Remove an overly broad entry, save, and test again. If the issue concerns a message that needs human context, move that case to alert moderation rather than trying to encode every possible context in the word list.
Finally, check whether the dashboard’s current wording or scope differs from the instructions you followed. Streamlabs’ older dedicated filter article refers to alerts, widgets, and Recent Events; do not assume every current setting has exactly that scope without checking the dashboard. For a broader operational checklist on a continuous channel, see setting up a 24/7 YouTube stream with OBS on an AWS virtual machine, while keeping alert moderation as a separate concern.
When nobody can sit at the controls all night, a message that needs judgement can wait in an approval queue, but someone still needs to review it. StreamNeo removes the need to leave your own computer running for a prepared, continuous YouTube broadcast, so the operator can focus on the channel and its moderation process rather than babysitting playback.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Does the profanity filter block every unwanted alert message?
No. It applies word-based rules and should be treated as a useful layer, not a guarantee. Unusual spellings, context, or text outside the configured alert controls can still require human attention.
Does YouTube’s blocked-word list filter Streamlabs alerts?
YouTube’s live-chat moderation settings apply to live chat in YouTube Studio. Do not assume they change the text shown by a separate Streamlabs Alert Box overlay; configure each system for its own purpose.
Should I censor a word or disable the whole alert?
Censoring keeps the event visible while obscuring a matched word. Disabling the alert can hide a legitimate event along with its message, so choose the behaviour that fits the alert and verify it in the current dashboard.
When should I use alert moderation instead of a word list?
Use moderation when a person needs to judge context before an alert appears, or when the messages are too unpredictable for a practical list. It requires someone to review and approve queued alerts, so plan for times when the channel is unattended.