Tourist Translator: voice translation for talks and lectures
Say a phrase, tap Done and hear it in the other language. Long speech switches to lecture mode. 84 languages, paid in tokens per minute of speech.
Updated November 2, 20265 min read
Tourist Translator is for speaking with someone face to face when you don't share a language: at a hotel desk, in a taxi, at a pharmacy. It also follows a long talk, such as a lecture, a sermon or a guided tour, and shows the translation on screen as the person speaks. Open it from the Tools tab, from the Tourist quick action (long press on the app icon) or with Siri: "Tourist translator in VocaLingo".
How to translate a conversation
The screen has two blocks, one for each language. Each person speaks into their own block, and the translation appears and is read aloud in the other one.
Tap the language at the top of each block and pick the two languages. The arrows between the blocks swap them.
Tap Start in the block of the language you are about to speak.
Say the phrase and tap Done.
Your words appear in your block. The translation appears in the other block and plays automatically.
Hand the phone over or turn it around. The other person taps Start in their block and answers.
One phrase can be up to 90 seconds long. Under each translation there are buttons to play it again, copy it or clear the block.
Lecture mode for long speech
Lecture mode turns on by itself when someone speaks for longer than 8 seconds without a break. The translation then appears in parts while the person keeps talking, and the button changes to Stop. Tap it when the talk is over. How to prepare for a long talk, what to do with a weak connection and how to keep the text: How to translate a lecture, sermon or guided tour live.
In lecture mode the translation is shown as text only. It is read aloud at the end only if the speech lasted less than a minute.
One recording lasts up to 1 hour. After that the recording stops, and you tap the original text to continue.
Keep VocaLingo open on the screen. If you switch to another app or lock the phone, translation stops: "Translation stopped while the app was in the background. Tap the original to continue". The screen stays on by itself while VocaLingo is listening.
If the app hears nothing for 20 seconds, it shows: "No speech for 20 seconds. If someone is speaking, move the phone closer to them or the speaker".
Settings
The settings button is at the top right of the Tourist Translator screen.
Setting | What it does |
|---|---|
Translation model | Gemini 3.8 Flash is the default. Gemini Flash Lite is the fastest, but it only translates the most common languages (English, Russian, Spanish, French, German, Italian, Portuguese, Chinese, Japanese, Korean, Turkish); for other pairs Gemini 3.8 Flash is used automatically. GPT 6 Luna is the OpenAI option. |
Voice for each block | George, Sarah or Daniel. By default both blocks use a male voice. |
Autoplay translated speech | On by default. When it is off, the translation is only shown, and you tap play to hear it. |
Translation text size | S, M, L or XL for the lecture mode text. |
Show the original | Three lines of the original above the translation in lecture mode. Turn it off to give the whole screen to the translation. |
Which languages are supported
Tourist Translator supports 84 languages. All of them can be voiced except Khmer, Albanian and Zulu: for these three you get the translation as text. The full lists for each tool are in Supported languages.
How many tokens it uses
Tourist Translator is paid in tokens. Before the microphone starts, the app checks that your balance is above zero; if it is empty, you see "Insufficient tokens". Tokens are spent only on speech that was actually processed, not on the time the screen is open.
What is charged | Price |
|---|---|
Speech recognition | about 1.3 tokens per minute of speech (78 per hour) |
Translation | a fraction of a token per phrase |
Voiced translation | 10 tokens per 1,000 characters of translation |
Small amounts add up and are deducted once they reach a whole token, so Usage History shows whole tokens.
How much 2 hours cost
A 2-hour lecture where someone talks almost without pauses: recognition alone is about 156 tokens, and translation adds about as much again. In real lectures this comes to 1.5 to 2.7 tokens per minute, so plan for about 300 tokens. Lecture mode has no voiceover, so you don't pay for voice.
A conversation that lasts 2 hours costs much less, because people speak only part of that time. A short phrase with a voiced translation costs about 0.3 tokens, so 100 phrases come to about 30 tokens.
Prices for all tools are in Token pricing for each tool, and ways to top up are in How to buy tokens.
Where conversations are saved
Each conversation is saved to History as one entry with the number of translations in it. Tap History at the top of the screen to open it. Inside an entry, Discuss in chat sends the whole conversation to a new chat, where you can ask for a summary or check a phrase. Without Premium the last 3 conversations are kept; see History for details.
What it can't do
It needs the internet. Without a connection you see "Internet problem. Check your connection and try again."
It doesn't translate phone calls or calls in WhatsApp and Telegram. For calls, see Can VocaLingo translate a phone or WhatsApp call.
It needs microphone access. If the app shows "Failed to access the microphone.", follow Microphone or camera access denied.
Contact support
Describe the problem and we will reply by email. Add the file name or job ID if you have one.