{"id":86,"date":"2026-08-28T18:35:06","date_gmt":"2026-08-28T18:35:06","guid":{"rendered":"https:\/\/interaction-bot.com\/blog\/discord-voice-translator-bot\/"},"modified":"2026-08-28T18:35:06","modified_gmt":"2026-08-28T18:35:06","slug":"discord-voice-translator-bot","status":"publish","type":"post","link":"https:\/\/interaction-bot.com\/blog\/discord-voice-translator-bot\/","title":{"rendered":"Discord Voice Translator Bot: How to Translate Voice Messages and Voice Chats"},"content":{"rendered":"<p>A &#8220;Discord voice translator bot&#8221; is a bot that turns spoken audio into text, translates it into another language, and can read the translation back aloud. As of 2026 that works reliably for one thing: <em>voice messages<\/em>, the recorded audio clips people send in chat, which the bot transcribes and translates into a text reply. It also works as a live channel helper when the bot speaks translated text aloud via text-to-speech. What is still not solved is the thing the phrase sounds like it promises: true real-time simultaneous translation of a two-way voice call, where both people just talk and each hears the other in their own language. Treat any bot that claims that as an experiment, not a feature. Here is what a voice translator bot actually does, how to set one up, and where the honest limits are.<\/p>\n<h2>What a voice translator bot can and cannot do in 2026<\/h2>\n<p>The useful parts, in order of how solid they are:<\/p>\n<ul>\n<li><strong>Translate voice messages.<\/strong> Discord&#8217;s voice messages are audio attachments up to 20 minutes long, recorded on the mobile app and posted in a chat. A bot can read that audio, transcribe it with speech-to-text, translate the transcript, and post the translation as a text reply. This works today because the audio arrives as part of the message, so the bot never needs to sit in a voice channel to do it.<\/li>\n<li><strong>Read translations aloud.<\/strong> Once there is text, text-to-speech can speak it. A bot with access to a voice channel can broadcast a translated text message as speech, so someone who cannot or prefers not to read can follow in their own language.<\/li>\n<li><strong>Participate in a voice channel on the audio side.<\/strong> A bot can join a voice channel, transcribe what is said into a text feed, or speak typed messages into the room. This makes a voice channel work for members without a usable microphone or without the room&#8217;s spoken language.<\/li>\n<li><strong>NOT real-time interpretation of a live call.<\/strong> Translating two people talking over each other, live, with accents and background noise, while also producing a natural spoken translation, stacks three imperfect systems (speech recognition, machine translation, speech synthesis) in series. The latency and errors compound. As of 2026 no Discord bot reliably interprets a live voice conversation the way a human interpreter would. &#8220;Real-time voice translation&#8221; videos and listings exist, but what they demonstrate is usually triggered transcriptions or short read-aloud exchanges, not continuous call translation.<\/li>\n<\/ul>\n<h2>How voice translation works under the hood<\/h2>\n<p>Whatever the bot, the pipeline is the same four stages, and each one adds a little time and a little error:<\/p>\n<ol>\n<li><strong>Audio.<\/strong> The voice arrives as a recording attached to a message (a voice message) or as the bot&#8217;s audio capture from a voice channel it has joined.<\/li>\n<li><strong>Speech-to-text (STT).<\/strong> A speech recognition engine transcribes the audio into text and makes a best guess at the language. This is where accents, background noise, and overlapping speakers do the most damage: the rest of the chain can only work with the text it is given.<\/li>\n<li><strong>Machine translation.<\/strong> The transcript goes to a translation engine, the same machinery used for typed messages. Detection is a probability, not a fact, and short or slang-heavy text is easy to misclassify. Our explainer of <a href=\"https:\/\/interaction-bot.com\/blog\/how-automatic-translation-bots-work-discord\/\">how automatic translation bots work on Discord<\/a> covers detection confidence and the translation APIs underneath in more detail.<\/li>\n<li><strong>Output.<\/strong> The bot posts the translated text as a reply, an embed, or a webhook, and optionally feeds it to a text-to-speech engine so it is also spoken aloud in a voice channel.<\/li>\n<\/ol>\n<figure class=\"post-figure\"><img decoding=\"async\" alt=\"A studio condenser microphone photographed in close-up, representing the audio stage of voice translation\" src=\"https:\/\/images.pexels.com\/photos\/6190821\/pexels-photo-6190821.png?cs=srgb&amp;dl=pexels-arjen-klijs-3246016-6190821.jpg&amp;fm=jpg\" title=\"Credit: Pexels License, via https:\/\/www.pexels.com\/photo\/close-up-photo-of-a-condenser-microphone-6190821\/\" \/><figcaption>Credit: Pexels License, via https:\/\/www.pexels.com\/photo\/close-up-photo-of-a-condenser-microphone-6190821\/<\/figcaption><\/figure>\n<p>The speech-to-text and text-to-speech pair is exactly what a dedicated Discord voice bot demonstrates: transcribing what is said in a voice channel and synthesizing typed messages back into speech. Seasalt AI&#8217;s walkthrough of its SeaVoice bot shows both halves running on one server:<\/p>\n<p><iframe loading=\"lazy\" title=\"How to Use SeaVoice STT\/TTS Bot on Discord Voice Channel | Seasalt.ai Speech-to-Text Text-to-Speech\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/drOVk_bexFY?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/p>\n<p>That video shows the STT and TTS hardware of the pipeline. The translation stage in the middle is what a combined bot adds: Interaction Bot, the working example through this guide, bundles translation, speech-to-text, and text-to-speech in a single free bot, which is still rare. Its homepage states that it translates messages into over 100 languages and provides text-to-speech in voice channels. For contrast, most of the field covers one piece: SeaVoice, the bot shown in that walkthrough, is chiefly a speech-to-text and text-to-speech tool for voice channels; iTranslator, demonstrated in several setup videos, translates text messages but does not touch audio; and the wamellow bot focuses on text-to-speech in voice chats. If your goal is genuinely a &#8220;voice translator,&#8221; the one worth adding is the bot that does all three stages, because transcription alone or translation alone leaves your voice messages still sitting untranslated in the chat.<\/p>\n<h2>How to set up voice translation in a server, step by step<\/h2>\n<p><a href=\"https:\/\/doc.interaction-bot.com\/commands\/translation\">Interaction Bot&#8217;s own documentation<\/a> lists the pieces you will use: <code>\/translate &lt;message&gt; &lt;language&gt;<\/code>, <code>\/detect &lt;message&gt;<\/code>, <code>\/languages<\/code>, and voice commands <code>tts<\/code> and <code>transcribe<\/code>, all configured either by slash command or from the dashboard. The full menu-by-menu walkthrough is in our <a href=\"https:\/\/interaction-bot.com\/blog\/discord-translator-bot-setup-guide\/\">Discord translator bot setup guide<\/a>; here is the voice-specific sequence:<\/p>\n<ol>\n<li><strong>Invite the bot.<\/strong> Go to interaction-bot.com and click Add to Discord, choose your server, and authorize the request. You need the Manage Server permission or approval from someone who has it.<\/li>\n<li><strong>Grant only what the voice workflow needs.<\/strong> For translating typed messages and transcribing voice messages, the bot needs to View Channel, Read Message History, and Send Messages in the channels it works in. For TTS playback or live transcription inside a voice channel, it additionally needs Connect and Speak on that channel. Decline Administrator unless a bot explicitly requires it.<\/li>\n<li><strong>Configure translation rules for the text half.<\/strong> In the dashboard, open Translation and add Channel Translation rules: the channel to watch, the channel where the translation appears, and the target language. Flag Translation (a flag emoji asks for a language) is the lowest-noise option for channels where most people share one language. The mechanics of these rules, and of automatic per-channel translation more broadly, are covered in our guide on <a href=\"https:\/\/interaction-bot.com\/blog\/how-to-translate-messages-in-discord-automatically\/\">translating messages in Discord automatically<\/a>.<\/li>\n<li><strong>Use the voice commands.<\/strong> Have a member post or record a voice message, then run the bot&#8217;s transcription command on it to get text, and translate or read that text with the translation and TTS commands. In a live voice channel, the bot&#8217;s TTS command speaks a typed message aloud so a member can &#8220;talk&#8221; by typing, and transcription turns what others say into a readable feed.<\/li>\n<li><strong>Test before announcing.<\/strong> Post a real sentence in the source language (a full sentence gives detection much more to work with than a single word), check the transcription and the translation against the original, and confirm playback volume in the voice channel. Keep the original audio and text visible alongside the translation, because both transcription and translation can be imperfect.<\/li>\n<\/ol>\n<figure class=\"post-figure\"><img decoding=\"async\" alt=\"A studio microphone with headphones resting behind it, standing in for the voice-channel half of setup\" src=\"https:\/\/images.pexels.com\/photos\/3587475\/pexels-photo-3587475.jpeg?cs=srgb&amp;dl=pexels-chuck-3587475.jpg&amp;fm=jpg\" title=\"Credit: Photo by chuck, Pexels License, via https:\/\/www.pexels.com\/photo\/close-up-photo-of-headset-3587475\/\" \/><figcaption>Credit: Photo by chuck, Pexels License, via https:\/\/www.pexels.com\/photo\/close-up-photo-of-headset-3587475\/<\/figcaption><\/figure>\n<h2>What Discord itself offers, and where it stops<\/h2>\n<p>Discord has three built-in features that touch this subject, and none of them is a voice translator.<\/p>\n<p><strong>Voice messages.<\/strong> <a href=\"https:\/\/support.discord.com\/hc\/en-us\/articles\/13091096725527-Voice-Messages\">Discord&#8217;s official help article<\/a> describes voice messages as recordings sent through direct messages, group direct messages, or a server with voice messages enabled. Two facts matter here: they can be recorded only from the mobile app (press and hold the mic button), and they can be up to 20 minutes long. Anyone can listen on any device. Server admins can switch the feature on or off per role under Server Settings &gt; Roles, using the &#8220;Send Voice Messages&#8221; toggle. What Discord does <em>not<\/em> do is transcribe, caption, or translate those recordings. The audio just sits in the chat.<\/p>\n<p><strong>The <code>\/tts<\/code> command.<\/strong> <a href=\"https:\/\/support.discord.com\/hc\/en-us\/articles\/212517297-Text-to-Speech-101\">Discord&#8217;s built-in text-to-speech<\/a> reads a typed message aloud for eligible listeners through their own clients. As our guide to <a href=\"https:\/\/interaction-bot.com\/blog\/discord-text-to-speech-bot\/\">Discord text-to-speech bots<\/a> explains, it is a per-listener convenience tied to user and channel settings, not a shared voice-channel speaker, and it has no translation step. You still type in one language and hear one language.<\/p>\n<p><strong>Built-in text translation.<\/strong> Where Discord&#8217;s native translation is available for your account, it translates a single message you are reading, privately, into your app&#8217;s language. It is text-only, one reader at a time, on demand, and it never touches audio. In short, Discord&#8217;s built-in translation serves one reader privately; a translator bot serves the whole server and can be wired to voice.<\/p>\n<div class=\"table-wrap\"><table>\n<thead>\n<tr>\n<th>Task in 2026<\/th>\n<th>Discord built-in<\/th>\n<th>Bot with STT + translation + TTS<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Translate a recorded voice message<\/td>\n<td>Not possible<\/td>\n<td>Transcribes the audio, translates the text, posts the result<\/td>\n<\/tr>\n<tr>\n<td>Read a translation aloud to the room<\/td>\n<td><code>\/tts<\/code> reads one typed message to eligible listeners<\/td>\n<td>Bot speaks translated text into the voice channel<\/td>\n<\/tr>\n<tr>\n<td>Translate a live two-way voice call<\/td>\n<td>Not possible<\/td>\n<td>Still unreliable; not a dependable feature anywhere<\/td>\n<\/tr>\n<tr>\n<td>Translate text for everyone in a channel<\/td>\n<td>One reader, one message, privately<\/td>\n<td>Automatic rules per channel, role, or category<\/td>\n<\/tr>\n<\/tbody>\n<\/table><\/div>\n<h2>Limits and permissions, stated plainly<\/h2>\n<p><strong>Voice message length.<\/strong> The ceiling is Discord&#8217;s own: 20 minutes per voice message. Multilingual communities on Discord commonly use much shorter clips. For longer audio a bot may have its own upload or length caps, so check the bot&#8217;s documentation rather than assuming.<\/p>\n<p><strong>Language detection on audio is really text detection.<\/strong> The bot detects the language <em>after<\/em> transcription, which means recognition errors flow straight into detection and translation. A clean, complete sentence in one language is the easy case. Heavy accents, music, several people talking, or a word that exists in two languages all degrade it, and there is no way for the bot to &#8220;hear&#8221; tone or context that the transcript missed. The same is true of typed messages, which our automatic-translation article discusses in more depth.<\/p>\n<p><strong>Permissions: voice channel and microphone are not needed for voice-message translation.<\/strong> This is the confusion most people hit. When the job is translating a <em>voice message<\/em>, the audio is already attached to the message, so the bot processes the file directly and never joins a voice channel. Nobody needs a microphone connected to your server&#8217;s voice channels, and the bot does not need Connect or Speak. The only microphone involved is the sender&#8217;s phone while they record. You add Connect and Speak only when you want the additional features: TTS playback in a live voice channel, or the bot joining the channel to transcribe a conversation. For the transcription of a posted voice message, keep the permission grant to View Channel, Read Message History, and Send Messages.<\/p>\n<p><strong>Free and honest about quality.<\/strong> Translation bots differ in what &#8220;free&#8221; means. <a href=\"https:\/\/doc.interaction-bot.com\/faq\">Interaction Bot&#8217;s FAQ<\/a> is unusually direct about it: to keep translation unlimited and free, the bot uses less efficient translation models, and paying enables a higher-quality translation engine. That trade-off is worth knowing before you rely on a voice translation for something important, like a support or onboarding message. Keep the original text and audio visible, and use a paid engine or a fluent native speaker for anything where a wrong word costs money or trust.<\/p>\n<h2>The short answer<\/h2>\n<p>A Discord voice translator bot in 2026 is real and useful, but scope it correctly. It transcribes voice messages up to 20 minutes long, translates the transcript, and can speak the result aloud with text-to-speech, all through one bot like Interaction Bot that keeps the STT, translation, and TTS pieces under one dashboard. Live simultaneous translation of an actual voice call is not dependable yet, and nothing on the market should be sold as if it were. Set it up by inviting the bot, granting read, send, and (only if you want channel speech) Connect and Speak, configuring a channel translation rule, and testing with real sentences before you rely on it.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>A Discord voice translator bot transcribes voice messages with speech-to-text, translates the text, and reads results aloud with TTS. How it works, setup steps, and honest limits in 2026.<\/p>\n","protected":false},"author":1,"featured_media":87,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-86","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/posts\/86","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/comments?post=86"}],"version-history":[{"count":0,"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/posts\/86\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/media\/87"}],"wp:attachment":[{"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/media?parent=86"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/categories?post=86"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/interaction-bot.com\/blog\/wp-json\/wp\/v2\/tags?post=86"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}