Übersetz: Live Translator

Follow spoken conversations with live translations on your Android device. Übersetz is a live voice translator that turns speech into readable transcripts and translates it into your chosen language. Follow a meeting, understand a lecture, or read along with conversations while travelling. READ SPEECH AND TRANSLATIONS AS THEY ARRIVE Watch partial transcripts appear while someone speaks, followed by completed text. Switch between Heard and Translation to read the original speech or its translation. Choose the source language yourself, or use automatic source handling with supported live providers. If you only need transcription, select the same source and target language to skip unnecessary translation. Turn your device sideways for a large, immersive landscape view that is easy to follow across a room. Choose light, dark, or system appearance, with listening controls kept within easy reach. KEEP LISTENING IN THE BACKGROUND Start listening while Übersetz is open, then switch to another app or lock your device. Speech transcription and translation can continue through an Android foreground service while the system permits it. A persistent notification shows when Übersetz is listening and lets you stop the session. Android's microphone privacy indicator remains visible while the microphone is active. CHOOSE HOW TO TRANSLATE - Google On-device: ML Kit speech recognition and ML Kit Translation run locally on supported devices after the required models are downloaded. - OpenAI GPT Realtime Translate: sends microphone audio to OpenAI and returns live source and translated text. - Google Gemini Live Translate: sends microphone audio to Google and streams source and translated text. - Your own Ollama server: translates stable transcript text sent to your server. - OpenAI-compatible endpoints: translate stable transcript text at your configured endpoint, including compatible OpenRouter configurations. ON-DEVICE SUPPORT AND SETUP Google's on-device speech recognition requires a supported Android device and source language. Übersetz checks availability before you start a session and guides you through downloading the required speech and translation models. If on-device speech recognition is unavailable, Google On-device, Ollama, and OpenAI-compatible options are disabled because they depend on it. GPT Realtime and Gemini Live Translate remain available as cloud alternatives, subject to provider access and setup. PRIVACY BY CHOICE With Google On-device, transcription and translation stay on your device after the required models are installed. No account is required for supported on-device use. GPT Realtime and Gemini Live Translate receive microphone audio for transcription and translation. Ollama and OpenAI-compatible engines receive transcript text at the endpoint you configure. Provider credentials are encrypted using a non-exportable key held by Android Keystore. Übersetz has no ads or in-app tracking. External providers may require their own API key, network connection, account, or paid usage plan. REQUIREMENTS AND CURRENT LIMITS Requires Android 12 or later. Features depend on your device, selected language, installed Google components, provider availability, and network connection. Open Übersetz before starting a background microphone session. Force-stopping the app or Android terminating its foreground service ends listening. Transcripts are retained only while the app process remains active. Saved-session recovery, transcript sharing and export, and speaker detection are not included in this Android release.