DubStream Companion

For Microsoft Teams meetings

Hear every meeting in your language.

The companion runs next to Teams. It sends the meeting audio for live translation, plays the dubbed voice in English, Hindi or Chinese, and turns Teams down while the dub plays. Subtitles float over your screen.

Choose your computer below.

How it works

  1. The host installs the app and clicks Start hosting while in the Teams meeting. It captures what Teams plays plus the host's microphone.
  2. Share the listener link in the Teams chat.
  3. Everyone picks a language. People with the app hear the dub on their computer with Teams turned down. Anyone can also open the link on a phone with earbuds.

Use headphones. On speakers the dub can leak back into your microphone.

All downloads

Loading releases…

Opening the app the first time

These test builds are not yet signed by Apple or Microsoft, so your computer will warn you once.

macOS

  1. Open the .dmg and drag DubStream Companion to Applications.
  2. Open it. macOS says it can't verify the developer. Click Done.
  3. Go to System Settings → Privacy & Security, scroll down and click Open Anyway.
  4. Allow Microphone and System Audio Recording when asked. That lets it hear Teams.

Needs macOS 14.2 or later to capture only Teams. If macOS says the app is damaged, run xattr -dr com.apple.quarantine "/Applications/DubStream Companion.app" in Terminal.

Windows

  1. Run the installer. If you see Windows protected your PC, click More info → Run anyway.
  2. Allow microphone access when asked.
  3. To hear the dubbed voice at full volume during Teams calls, open Sound settings → More sound settings → Communications and choose Do nothing.

Windows 10 (2004) or 11. On Windows the host captures everything playing on the device Teams uses, so the host should not play the dub on that same device.

Privacy

While hosting, the meeting audio and your microphone are streamed to DubStream for translation. Tell the people in your meeting before you start, as you would for a recording. Listeners only receive audio and text; they send nothing.