Home / Discord live captions

Discord Live Captions & AI Voice Chat Translation

Discord is built around voice — gaming squads on a raid, study groups in a stage channel, AMA sessions on a community server, and Town Hall events with hundreds of listeners. But voice-only conversations are hard to follow when teammates have heavy accents, when background noise drowns the channel, or when half your friends speak a different language. Discord live captions via Live Subtitles fix this: you get real-time AI subtitles for every voice channel and stage event, with no Discord bot to invite and no server admin permission to ask for.

How to add live captions to Discord in 3 steps

1. Download Live Subtitles from the Microsoft Store and launch it alongside Discord.
2. Select "System Audio" as the source — it captures Discord voice chat without any microphone or bot setup.
3. Join any voice channel, stage, or event — captions appear instantly in a floating overlay that stays on top of your game.

Built for gaming squads and late-night raids

If you have ever been deep in a ranked match and missed a critical callout because your headset volume was masked by gunfire, Discord live captions are for you. Live Subtitles displays a real-time transcription of everything said in your voice channel, in a floating overlay that you can dock anywhere on screen. It even stays on top of fullscreen and borderless games, so you do not need to alt-tab out to read what your team is saying.

Common gaming use cases:

• Raid callouts in MMOs and tactical shooters where missing a single instruction wipes the team.
• International squads where some teammates speak English as a second language.
• Streamers who want a clean text record of squad chatter for editing highlights later.
• Players with hearing difficulty who want to follow strategy without cranking the headset.

Works with stage channels and server-wide events

Discord stage channels — used for AMAs, podcast-style discussions, conferences, and large community events — broadcast one or more speakers to a passive audience that can number in the hundreds. Stage events often run long, cover dense technical material, and host speakers from many countries. Live Subtitles captions stage audio just like any other voice channel, so you can follow a 90-minute panel without losing focus when an accent or fast speaker pace pushes you out of the conversation.

Because Live Subtitles reads system audio, it does not care how the audience is structured: speakers on the stage, listeners in the audience, or moderators handling Q&A — every voice that comes through your headphones is captioned in real time.

Translation for multilingual servers and language-learning communities

Discord has become a hub for language-learning communities — Japanese study servers, Korean K-pop fan groups, Spanish conversation rooms, German tabletop circles. The members want to practice the target language but also want to understand each other. Live Subtitles dual subtitle mode is built for this: the original speech appears on the top line, your translation language appears below — both at the same time. You can listen for pronunciation while reading the meaning, then switch the translation off entirely once your comprehension catches up.

Supported: 50+ recognition languages and 50+ translation target languages including Japanese, Korean, Chinese, Spanish, French, German, Italian, Portuguese, Russian, Arabic, Hindi, Polish, Dutch, Turkish, and Ukrainian.

No bot, no server admin, no Discord plugin

Live Subtitles is a standalone Windows app. It does not join your voice channel as a bot user, does not appear in the Discord member list, and does not need any manage-server or manage-roles permission on the server. This matters: many community servers ban third-party voice bots outright because of privacy and trust concerns. Live Subtitles sidesteps that completely — it is just an audio reader on your own PC, captioning what you can already hear, and other members of the call have no way to know you are using it.

Go Live, screen share and watch-together

Discord is not only voice chat. Go Live streams a game or a window to the channel, screen share carries the audio of whatever is playing, and watch-together sessions run video for the whole group. All of it arrives at your machine as ordinary system audio, which means it is all captioned by the same setup with nothing to change.

That covers a case bots handle badly: a friend sharing a foreign-language video, a tutorial, or a game with heavy story dialogue. A bot receives participants' microphone streams, so shared media audio is largely invisible to it. An overlay reading system audio captions the shared content and the people talking over it alike.

Discord has no native live captions

This surprises people, because Discord does have an Accessibility tab. What lives there is reduced motion and the /tts text-to-speech command — which reads typed messages aloud, the opposite direction from captioning. As of 2026 there is still no Discord setting that transcribes voice channels, despite live captioning being one of the longest-standing requests on Discord's own feedback board.

So every option is third-party, and they split into two families with very different trade-offs.

Transcription bots vs. a desktop overlay

Bots such as Scripty join the voice channel as a member and post transcripts into a text channel. Their real advantage is speaker labels: because the bot receives each participant's audio stream separately, it can attribute every line to a person. That is genuinely useful for a meeting-style call you want a record of.

The costs are equally real. A bot needs someone with Manage Server permission to invite it, it transcribes everyone in the channel into a log that persists, and plenty of community servers ban third-party voice bots outright over exactly that. It also puts the transcript in a text channel you have to look away from the game to read.

Desktop overlays like Live Subtitles work the other way round. They caption the mixed audio coming out of your headphones, so there are no speaker labels — but there is no bot in the member list, no permission to request, no server-side log, and the text appears on top of your game instead of in a channel you have to alt-tab to. Windows also ships a basic version of this idea in Live Captions (Win+Ctrl+L), which is free but English-first and has no translation.

The short version: pick a bot if you want an attributed written record of a call, and a desktop overlay if you want to follow the conversation in real time while doing something else.

Privacy in a shared voice channel

Worth stating plainly, because the two approaches are not equivalent here. A desktop overlay captions audio that has already been played to you, on your own machine, and nothing leaves it — there is no transcript stored on a server, nothing posted to a channel, and no way for other members to see that you are using it. That is why it is viable on servers where voice bots are banned.

A transcription bot is a different proposition: it creates a persistent record of what everyone said, usually visible to the whole channel. That may be exactly what you want for a study group or a team call, and exactly what you should not deploy on a casual server without telling people. Norms about recording vary a lot between communities — and in some jurisdictions so does the law — so the honest default is to say what you are running before you run it.

If captions don't appear

Discord's audio routing causes nearly all of these:

• The source is set to Microphone. That captions you, not the channel. For voice chat you want System Audio.
• Discord is sending audio to a different device. Check Discord's Voice & Video settings — the output device there must match what the app is capturing. A headset selected in Discord while the app listens to desktop speakers produces silence.
• Everyone is on push-to-talk. Nothing is wrong; there is simply no audio between transmissions. Captions resume as people speak.
• The overlay is behind a fullscreen game. Game Mode (Ctrl+Shift+L) makes it click-through and keeps it on top; a second display avoids the conflict altogether.

Download Live Subtitles — Free Trial

Related platform guides