Home / Android live captions
Live Captions & Dual Subtitles on Android
Live Subtitles puts AI-powered live captions and dual-language subtitles in a floating overlay on your Android phone — over YouTube, a show, a stream, a game, or any other app. Switch to the microphone and it captions and translates a conversation happening in the room, no app on the other end required.
Why Android's built-in captions aren't quite this
Android has shipped on-device live captioning since Android 10, and it's genuinely useful — but it solves a narrower problem:
- No translation. Android's system caption feature transcribes audio in the language it hears. It doesn't translate into another language or show two lines at once.
- No microphone conversation mode. It captions what your phone plays, not what's being said in the room by someone standing in front of you.
- Nothing persists. There's no dual-subtitle view to follow a foreign-language video while still reading the original line.
Live Subtitles is built specifically for the translation case: two capture modes (system audio and microphone), 50+ languages, and both the original and translated line shown together.
What Live Subtitles adds on Android
A floating overlay over any app
Once the overlay permission is granted, captions sit on top of whatever you're doing — switch between YouTube, a stream, or a game, and the subtitles stay visible, draggable, and resizable.
Two ways to capture audio
System Audio reads whatever the phone is playing — videos, games, podcasts, livestreams — through Android's screen/audio capture API. It can't read calls or voice chats: Android keeps that audio separate from media playback, by design, regardless of whether it's a regular phone call or a call made through an app like WhatsApp, Zoom, Google Meet, Microsoft Teams, or Discord. Microphone mode covers that case instead — it listens through the phone's mic, for a call on speaker, an in-person conversation, a lecture, or a guide speaking a language you don't.
Dual subtitles, 50+ languages
See the original spoken line and your chosen translation together on screen, rather than losing the source wording the moment it's translated. Useful both for following the actual meaning and for picking up the language as you go.
Customizable overlay
Font size, text color for the original and translated lines, background opacity, and on-screen position are all adjustable, and the overlay remembers where you left it between sessions.
Setup: a few steps, once
- Install Live Subtitles from Google Play. Opening it starts the free trial — no account required.
- Grant "Display over other apps." The app opens this Android settings screen for you the first time; turn the toggle on and go back.
- Pick a source. System Audio for anything the phone plays, or Microphone for a live conversation. System Audio asks Android's own one-time capture-consent prompt each session — that's an Android requirement for every app that reads screen/audio output, not something Live Subtitles adds.
- Choose languages and start. Pick the spoken language (or auto-detect) and your translation language, turn on dual subtitles, and the overlay appears.
Android's own captions vs Live Subtitles
| Feature | Android's built-in live caption | Live Subtitles |
|---|---|---|
| Translation into another language | No — same-language transcription only | Yes, 50+ languages |
| Dual-language display | No | Original + translation together |
| Captions a live, in-person conversation | Not built in | Microphone mode |
| Overlay works over any app | Yes | Yes |
| Customize font, color, opacity, position | Limited | Full control |
| Internet required | No — on-device | Yes — cloud speech and translation |
Real-world use cases
Foreign-language videos and vlogs
Watch a YouTube video or stream in a language you're learning or don't speak, with the original audio's text and a translation on screen together, instead of relying on auto-generated captions baked into the platform.
Livestreams, podcasts, and recorded lectures
Caption and translate a Twitch or YouTube Live stream, a podcast episode, or a recorded lecture or course video straight from System Audio, without the platform's own auto-captions or a lack of translation getting in the way.
Travel, calls, and in-person conversations
Switch to Microphone mode to get a live translated transcript of what someone nearby is saying — ordering food, asking directions, following a guide, or taking a call on speaker — without handing your phone back and forth through a separate translator app.
Language learning on the go
Dual subtitles let you read the original line while watching the translation land, which is a common technique for picking up vocabulary and sentence structure from native content.
Tips for better results
- Grant the overlay permission once, up front — the app will prompt for it automatically, but granting it manually in Settings first avoids an extra step mid-session.
- Set the source language explicitly when you know it; auto-detect takes a little longer and can confuse closely related languages.
- Keep the app out of aggressive battery-optimization lists on phones that kill background foreground services aggressively (common on some manufacturers' Android skins), so a session doesn't get cut off.
- Use Microphone mode for DRM-protected streaming apps instead of System Audio, since their audio can't be captured directly.
Pricing and free trial
Live Subtitles is free to install, and every new install starts with a free trial balance you can use immediately — no credit card needed. After the trial, you can subscribe or buy one-time minute packs directly in the app; current pricing and plans are shown there and on the Google Play listing. The same account also covers the Windows, macOS, and iOS versions.
Get it on Google PlayAlso available for Windows and iOS / macOS.
FAQ
How do I let the app show subtitles over other apps?
Live Subtitles opens the "Display over other apps" permission screen for you the first time you start a session — toggle it on. You can also set it manually under Android Settings > Apps > Live Subtitles > Display over other apps (naming varies slightly by phone brand).
Which Android versions and phones are supported?
Android 10 (API level 29) and newer, on any manufacturer's phone — the overlay and audio-capture APIs it uses are standard Android, not brand-specific.
Do I need an internet connection?
Yes. Speech recognition and translation run in the cloud, so a stable connection is required. There's no offline mode.
Does it work with Netflix, Prime Video, or other streaming apps?
System Audio mode captures most apps — YouTube, browsers, games, podcasts, livestreams. DRM-protected audio from services like Netflix, Prime Video, or Disney+ is muted by Android for any capture, by design. Switch to Microphone mode and point the phone at the speaker to still get captions there.
Does it translate WhatsApp, Zoom, Teams, or regular phone calls?
Not through System Audio. Android keeps call and voice-chat audio separate from media playback, so System Audio mode can't read it — whether it's a regular phone call or a call through WhatsApp, Telegram, Zoom, Google Meet, Microsoft Teams, or Discord. Switch to Microphone mode and let it listen through the phone's mic instead.
Can it caption a live, in-person conversation?
Yes — Microphone mode transcribes and translates speech happening in the room, useful for travel, calls on speaker, or talking with someone who doesn't share your language.
How many languages does it support?
50+ languages for translation, with dual subtitles showing the original and translated line together.
Is there a free trial?
Yes. New installs start with a free trial balance, no credit card required. Subscriptions and one-time minute packs are available afterward, in the app and on the Google Play listing.
Why does Android show a notification while I use it?
Capturing audio in the background requires a foreground service, and Android shows a persistent notification whenever one runs. It's a standard disclosure, not a sign of extra recording.
Does it tell different speakers apart?
No. It shows the original speech and its translation, but doesn't label or distinguish speakers.