SYSTEM AUDIO

How to translate system audio on Windows in real time

Translate whatever Windows is playing. All of it.

The short answer

Install Voxis, pick Video & Game, choose the language you want to hear and press Start. Voxis captures everything Windows is playing, in any app, and says it in your language a few seconds later, with live captions. There is no virtual cable, no audio driver and no file to upload.

  1. Install Voxis from the Microsoft Store, open it and sign in.
  2. Pick Video & Game and choose the language you want to hear.
  3. Press Start, then play anything: a browser tab, a video player, a game or a call.

The original sound drops while someone speaks and the translation is spoken over it. The sections below explain how the capture works and when you need two-way Meeting mode.

VoxisLive translates sound from your Windows PC — from any application — into your language in real time and speaks it back aloud. No file uploads, no pasted links: it captures the entire system audio mix as it plays. On Linux, Voxis does the same thing without a driver: it creates its own PipeWire virtual sink on demand and routes the app you are translating into it. Install it from the Snap Store with sudo snap install voxis.

On a Mac? See how to translate the audio playing on your Mac, or the overview of Voxis for Mac.

Any app's audio — not just one file

Most translation tools make you commit to a source up front: a file, a link, a tab. VoxisLive listens to the Windows system audio reaching your speakers, which means it works with any application — browsers playing foreign news, desktop video players, games with foreign voice acting, conferencing windows, podcast apps.

No per-source configuration: start VoxisLive, select your target language from 79, and play whatever you want. The translation follows you seamlessly across apps because everything is captured at the system level. Games are a good example: how to translate game voice chat in real time on Windows.

Driverless capture, explained

VoxisLive uses WASAPI process-loopback — built-in Windows functionality — to capture already-playing audio. No virtual audio driver, no VB-CABLE, no routing utility: on Windows 10 and 11 it works immediately after installation. Only two-way Meeting mode needs VB-CABLE, to carry your translated voice into the call. Capture excludes VoxisLive's own output, preventing re-translation loops.

Speech streams to a simultaneous interpreter model that recognizes, translates and re-speaks in one low-latency pass, staying just a few seconds behind the original talker.

One-way or two-way

For listening, Video & Game mode ducks the original audio so the translation sits clearly on top. For conversations, Meeting mode also translates your own speech into the other side's language through a virtual microphone — crucially, without adding any bot participant to the call. For a call, follow how to translate a Microsoft Teams call in both directions. For a voice channel, follow how to translate Discord voice chat in real time.

Spoken output, with a transcript when you need it

The result is a natural voice speaking your language — not screen captions. A live bilingual transcript is still maintained and can be exported as TXT, SRT or VTT, with searchable history.

VoxisLive ships primarily through the Microsoft Store, with a paid Bring Your Own Key unlock inside the app for using your own Gemini or Qwen key — see GitHub for how it works. See pricing to compare.

FAQ

Common questions

01Can VoxisLive translate audio from any app on Windows?
Yes. It captures the whole Windows system audio mix — browsers, desktop players, games, conferencing apps, media players. You don't select a file or a single app; it translates whatever is currently playing.
02Do I need a virtual audio cable to translate system sound?
No — not to capture your PC audio. VoxisLive uses driverless WASAPI process-loopback built into Windows 10 and 11 — no VB-CABLE, no virtual driver, no meeting bot. It also excludes its own output from capture. Two-way Meeting mode is the exception: on Windows it sends your translated voice through a virtual audio cable such as VB-CABLE and will not start without one; on Linux the app creates its own virtual microphone.
03Does it show subtitles or speak the translation?
It speaks the translation aloud in a natural voice — speech-to-speech, not subtitles. A live transcript is available and exportable as TXT, SRT or VTT.
04How many languages can it translate into?
79 target languages, using a native simultaneous interpreter model that translates while the speaker talks, staying only a few seconds behind.

Keep reading

WASAPI LoopbackTranslate any Windows system audio in real time without a virtual audio cable.How it worksVoxisLive captures system audio directly through WASAPI loopback, detects speech on-device, translates it with a…VideoVoxisLive speaks a live translation of any video playing on Windows — Netflix, YouTube, VLC, local files.
Free to start · 10 free minutes every day

Hear every language, in real time.

Runs on Windows 10 and 11, Mac (Apple silicon, macOS 15 or later) and Linux — driverless capture, no setup ritual, no bot in your call.

Get it on Microsoft Store Linux · Snap Store View on GitHub