Donation text-to-speech (TTS) is usually configured on the tip alert in the service that handles your alerts, then played through that service’s alert widget or overlay in your streaming scene. You need both parts configured: turning on TTS in a dashboard will not make messages audible if the relevant alert source is missing or its audio is not reaching the stream.
The exact menus vary by provider. This guide uses Streamlabs’ documented route as a concrete example, then notes where a StreamElements setup differs; check the current instructions for whichever service and tipping page you use.
How donation TTS reaches the stream
A viewer sends a tip (often labelled a donation in conversation) with a message. The tip service passes the event to an alert widget, which displays an alert and, if TTS is enabled for that event, reads the message aloud. Your streaming software captures the widget’s visual and audio output and sends it as part of the live programme.
That chain has several points where setup can fail. The tip page must be connected to the alert service; TTS must be switched on for the tip event rather than just another event such as a follow; and the matching alert source must be active in the scene you are streaming. Finally, the speech must be audible in the actual programme mix, not only in your headphones or a dashboard preview.
Streamlabs’ Text-to-Speech setup article says viewers hear a message live when TTS is enabled and they send a tip. Its procedure requires an Alert Box source in the stream. Treat that as a description of Streamlabs’ workflow, not a rule that every alert provider uses the same labels or controls.
If you already have a tipping page and provider, start there: use its matching alert service so the payment, message and alert event are connected. If you are building the setup afresh, compare whether the provider works with your streaming software and the page viewers will actually use. For a scheduled channel built around recorded programming, an alert is a separate layer from the video loop; planning the programme separately, as in this guide to streaming prerecorded videos continuously to a scheduled YouTube event, can help you avoid mixing up the content feed and its interactive alerts.
Add the alert or overlay source
In Streamlabs’ documented route, add an Alert Box source to your streaming scene, then configure it in the Streamlabs dashboard. In OBS or other streaming software, Streamlabs documents copying the Alert Box widget URL and adding it as a Browser Source. Streamlabs Desktop and its OBS plugin also have a native Alert Box source. Its Alert Box setup guide describes those options.
A Browser Source is a web page displayed inside a scene. It can show the alert without placing a separate browser window over your desktop capture. Add it to the scene used for the broadcast and check that it is visible above the video or other scene layers. If it is hidden behind an opaque source, the alert may be firing while remaining invisible to viewers.
Keep the widget source associated with the service that receives tips. If you have several scenes, decide which ones should show alerts and add or share the source accordingly. A source in a starting-soon scene does not necessarily appear in the scene you use for the main broadcast. This matters for a long-running channel where you may switch between a continuous video, a welcome screen and a break scene.
Other providers may work differently. In StreamElements, the documented flow is to create or edit an overlay, add an AlertBox widget, configure the tip alert, and add the overlay URL as a Browser Source in OBS when the overlay is separate. Check its alert configuration instructions for the current route. Do not assume that a Streamlabs menu name, source type or setting exists in another service.
Before moving on, confirm the source belongs to the service and account linked to the tipping page. If the dashboard test appears in one overlay but the scene contains a different overlay URL, you are testing the wrong connection. The OBS and FFmpeg scheduling comparison can help you think through the role of scene-based software in a scheduled setup; it does not remove the need to configure the alert source itself.
Open the tip alert’s TTS settings
For Streamlabs, log in to Streamlabs.com and open Alert Box. Select Tips (or Bits, if you are configuring that event) and open Text-to-Speech. Set the available TTS controls for that specific alert, then toggle TTS on. The service’s interface can change, so if the labels differ, use Streamlabs’ current support article rather than applying another provider’s steps by guesswork.
This event-level choice matters. A provider may offer separate alert settings for tips, follows, subscriptions or other events. Enabling speech globally, if a service offers such a setting, does not necessarily switch it on for tips. Confirm you are editing the alert viewers trigger when they send a tip through your page.
In StreamElements, open the tip alert settings for the AlertBox widget and look for its TTS controls. Inspect alert variations as well as the base alert: a variation can override the basic configuration. StreamElements’ setup article focuses on Twitch alerts and says the process is broadly similar elsewhere, but that is not a reason to promise identical settings across platforms. Check the instructions for your platform and overlay.
Also check that tips are configured and connected to the account. An alert widget alone does not create a working payment route. StreamElements’ tip settings documentation covers provider configuration and related settings; availability may depend on the provider and your circumstances. Do not promise viewers a particular payment method until you have confirmed it is offered to you and works on your page.
Choose a voice and minimum threshold
Choose a voice that fits the channel and is easy to understand over its usual sound. A devotional channel with quiet bhajans, a study stream and a local news loop have different listening contexts. Listen to a test message with your normal programme audio playing. A voice that sounds clear in an empty preview may be difficult to hear once music, commentary or other alerts are present.
Streamlabs’ guide describes voice choices and a minimum tip amount. The threshold limits which tips trigger speech; it does not have to match the smallest amount your page accepts. A lower threshold can make more messages audible, but can also interrupt the programme more often and increase the amount of unscripted text you must moderate. A higher threshold means fewer alerts speak, but some viewers’ messages will not be read aloud.
Set the threshold according to how often you are willing to pause the listening experience and how much review you can do. For example, a quiet overnight ambience channel may want spoken alerts to be uncommon, while a live community programme may choose to read more messages. Those are editorial choices, not settings that guarantee a particular audience response. If your provider does not offer a threshold, do not look for a universal substitute; use the controls it documents and consider whether TTS is suitable for that alert setup.
Voice lists and choices can change. Streamlabs’ 2025 guide stated that it offered over 50 voices at that time; this is a vendor feature count, not an independently verified market figure or a promise that every voice remains available. Choose by listening and checking language support rather than relying on a count. If you serve listeners in more than one language, test names and commonly used words before publishing the setting.
You may also be able to change the alert’s duration, text display or sound. Keep those decisions separate from TTS: a visual message can be useful when speech is muted or hard to understand, and an alert sound can compete with the spoken content. If the channel runs continuously, such as a 24/7 Arabic instrumental music stream, consider whether a spoken donation message should interrupt the same way at every hour.
Set spam and profanity controls
Donor messages are user-submitted text, so a voice setting alone is not a moderation plan. Before enabling speech, review the provider’s profanity, blocked-word and spam controls. Consider what should happen when a message is caught: some systems can replace or hide words, suppress the alert, or hold a tip for approval. The available choices differ by service.
Streamlabs documents a spam filter for TTS and profanity-filter options in its alert settings, including default and custom filtering. Its guide says the spam filter is off by default and suggests beginning at Low or Medium. Use that as a starting point from the vendor’s instructions, not as a universal setting. Test whether the chosen filter catches repetitive or abusive messages without silencing ordinary names and phrases your viewers are likely to use.
StreamElements documents standard and custom profanity protection, message filters, repeated-character TTS anti-spam and optional manual tip approval. It also states that filtered tip events can remain in the activity feed without triggering an alert, and that tip moderation settings do not affect fund transfers. Read the current moderation documentation for the behaviour of its controls rather than assuming that a filter has the same effect in another service.
Manual review can be useful if messages are read aloud on a family, school or community channel, or if you cannot monitor the alert feed continuously. It introduces a delay and requires someone to review messages; that may not suit a channel run alone overnight. If approval is unavailable, make the threshold and filtering choices with the risk of immediate speech in mind. Avoid implying that any filter will catch every offensive or disruptive message.
Check how moderation interacts with your chosen threshold and alert variations. A base tip alert may be filtered while a variation has a different configuration. If multiple people can edit the dashboard, agree who can change blocked words and approval settings. A simple record of the settings and the date you last tested them makes it easier to investigate a surprise alert later.
Save and test the alert
Save the dashboard changes, then test in two places: first in the service’s preview or test controls, and then through the streaming application. Streamlabs documents using its test dropdown, Live Preview or software test widgets to inspect an alert. A dashboard preview helps confirm that the alert is configured; it does not prove that viewers can hear it in the live programme.
In the streaming application, trigger a test while monitoring the programme output. Confirm that the alert appears in the intended scene, that the spoken message is audible in the stream mix, and that it does not drown out the underlying audio. If you use OBS, check Browser Source audio and the selected audio device or monitoring route when speech is missing. StreamElements’ audio troubleshooting guidance covers checks for alert audio and muted or paused alerts; use the troubleshooting page for your provider if its controls differ.
Test a message with punctuation, a long name or a phrase in the language your audience uses. Do not use a real donation as a test unless you understand the payment and refund implications. Use the provider’s test tool where possible, and keep the test text safe for public playback. Listen to the broadcast output on a separate device if you can, since local monitoring can sound different from the stream mix.
If the visual alert appears but speech does not, check that TTS is enabled for the tip event, the alert source is the correct one, and browser audio is included in the stream. If neither appears, verify the overlay URL or source, confirm the source is in the active scene, save settings and reload the overlay. A dashboard preview and the live scene are separate checks, not interchangeable ones.
If the test behaves as expected, keep a note of the selected voice, threshold and moderation settings. Revisit them after a service interface update or a change to the tipping page. If the channel’s main requirement is to keep a recorded programme running while your own computer is switched off, that is a different operational problem from TTS alerts: StreamNeo turns an uploaded video into a 24/7 YouTube live stream, so it removes the need to keep your computer running for that continuous video. It does not replace the alert service configuration described here.
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 donation TTS work if I only switch it on in the dashboard?
Not necessarily. The relevant alert source or overlay must also be included in the scene, and its audio must reach the programme output. Test in both the dashboard and streaming application before going live.
Do Streamlabs and StreamElements use the same settings?
No. They provide different workflows and may use different labels, controls and overlay arrangements. Follow the current instructions for the service connected to your tipping page, and do not assume that a setting in one service exists in another.
Can every tip message be read aloud safely?
No filter can be assumed to catch every message, so moderation is part of setup. Use the available spam and profanity controls, consider approval where offered, and test how the chosen controls treat ordinary messages as well as unwanted ones.
Do I need special hardware for donation TTS?
The documented workflows rely on service settings and an alert source in streaming software; they do not call for a special microphone, capture card or stream controller. You still need to confirm your software’s audio routing so the alert speech is included in the broadcast.