Linguago: Offline Translate

Linguago translates spoken language on your phone with the internet switched off. Speak a phrase, see it translated, and hear it read back. The AI model runs entirely on your device, so nothing you say is ever uploaded, stored on a server, or seen by anyone else. Built for the moments connectivity fails you: a foreign airport with no roaming, a rural clinic, a market stall three streets past the last usable signal. HOW IT WORKS Tap the microphone and speak. Linguago transcribes what you said and translates it in one step, then reads the result aloud in the other language. Tap the text box to type instead useful in noisy places, or to correct a word it misheard. Swap the direction with a single tap to hold a back-and-forth conversation. GENUINELY PRIVATE • No account, no sign-in, no email address • Your recordings are deleted the moment they have been read • No analytics, no crash reporting, no advertising • After first-time setup, the app makes no network requests at all This is not a claim about a privacy policy. There is no server for your voice to reach. BEFORE YOU DOWNLOAD Linguago needs a one-time 2.6 GB download to install its translation model. This takes roughly 5 to 15 minutes on Wi-Fi. Everything works offline afterwards. The model also needs a device with at least 6 GB of memory. On older devices, Linguago tells you plainly rather than crashing. LANGUAGES English, French, Spanish, German, Italian, Portuguese, Dutch, Russian, Arabic, Hindi, Chinese, Japanese, Korean, Turkish and Polish. BETA Linguago is early software. Speech recognition can mishear you, and translation quality varies between languages , English, French and Chinese have had the most testing. If a translation looks wrong, type the phrase instead. The text box is more reliable than the microphone, and the app says so where it matters. We would rather tell you this up front than have you discover it. OPEN SOURCE Linguago is open source under the Apache 2.0 licence. You can read every line, including exactly how audio is handled: github.com/Darkbeast-glitch/linguago