How it works
Speak your language. Hear theirs in yours.
Speak Tongues listens as people talk, turns each sentence into text, translates it and reads it to whoever needs it. Here's what that looks like in each kind of conversation.
One-on-one
For a conversation with one other person: a shopkeeper, a taxi driver, a new neighbor, your partner's grandmother.
- Pick their language, or choose Auto and Speak Tongues works it out from what they say.
- Talk normally. It tells who is speaking which language by itself, so there's nothing to press between turns.
- Their words appear on your screen in your language, and can be read into your earbuds.
- Your words come back in their language. Hear it quietly in your earbuds and say it yourself, or tap Say it out loud to play it on your phone's speaker for them.
- New alphabet? When their language is written in a different script from yours, your phrase is spelled out so you can read it aloud.
Group conversations
For a dinner table, a family visit, a tour or a meeting where people speak different languages. Everyone speaks their own and hears everyone else in theirs.
- Start a group in the app. It shows a QR code and a short code, like K7QM-4XPD, and a link you can send.
- Guests join in their browser. They scan the code with their camera or open the link, type their first name, pick their language and tap Join. No app, no account, no email.
- Each person uses their own phone with earbuds. Everyone reads the conversation in their own language and can tap a line to see what was actually said.
- Nothing is skipped. When people talk fast, lines wait their turn and play in order. Anyone can skip to the newest line if they'd rather catch up.
- Joining from somewhere else? Guests who aren't in the room tick a box when they join and hear every line, even the ones in their own language.
- Loud room? Hold to talk, so your phone only listens while you hold the button.
- At the end, the host keeps the transcript in their history, and each guest can download their own copy for 30 minutes.
The host pays for everyone, and each conversation has a spending limit.
Signs and menus
Point your camera at a sign, a menu, a label or a timetable, in the app or right here on the website.
- Each line comes back in your language, next to the original.
- How to say it is spelled out when the original uses an alphabet you don't read.
- The photo is discarded once it's translated. The text stays in your history for 30 days.
- About a cent or two a photo, on the same tab as your conversations.
Why earbuds?
Private listening
Each person hears the translations meant for them, in their own ear. In a group, the room doesn't fill up with overlapping voices.
A cleaner microphone
When translations play in your ears, your phone's microphone doesn't pick them up and mistake them for you. While one plays, your microphone pauses; start talking and it stops to listen, then plays again when you're done.
Any pair works
Bluetooth or wired, earbuds or headphones. Without them, you can still read every line on screen, and in one-on-one play your phrases on the speaker.
Privacy, in plain words
Your voice isn't kept
Audio goes straight to the transcription service and is never stored. Only the text of what was said is kept.
Your history is yours
Conversations and signs stay in your history for 30 days, then they're deleted automatically. Delete any of them sooner, or your whole account, in the app or on your account page.
Guests know before they join
Before joining a group, each guest sees, in their own language, that their speech is transcribed and translated by AI and that the host keeps the transcript. Joining is agreeing; anyone can leave at any time.
No ads, no tracking
We don't show ads, track you, or sell or share your information. The AI work is done by OpenAI, which doesn't train its models on it. Your email is used only to sign you in and to answer you.
Ready when you are.
Free to download. Pay only for what you use.