Loquora 0.11.1: 58 Languages and Faster Voice Translation

Over the past week, more than the language count has changed in Loquora. We also updated the flow of a conversation: the app can start translating a long reply sooner and is more careful about choosing a point where a complete part can already be passed to the other person.

These changes are included in version 0.11.1 for Windows and macOS.

58 languages for live conversations

The previous language list was much shorter. You can now manually choose any of 58 languages for your own speech and for the other person's reply. The list covers widely used European languages as well as Arabic, Armenian, Belarusian, Georgian, Hindi, Kazakh, Korean, Mandarin Chinese, Persian, Thai, Ukrainian, Urdu and many others.

Speech goes through the complete process in the selected language: Loquora recognizes what was said, translates it and produces spoken output. Quality can vary by language pair, pronunciation and call conditions, so it is worth testing the exact combination you plan to use before an important conversation.

The complete list and practical setup advice are available in our language selection guide.

When to use Auto Detect

If people switch languages during a conversation, you do not always need to change the source language yourself. Auto Detect can recognize and switch among ten common languages: English, Spanish, French, German, Hindi, Russian, Portuguese, Japanese, Italian and Dutch.

The other languages are still available, but they must be selected manually. This distinction matters: 58 is the number of languages you can choose, not the number covered by automatic detection.

When you already know which language will be spoken, choosing it directly is usually more predictable. It removes uncertainty at the beginning of a short phrase and in cases where words sound similar across languages.

Regional options improve recognition

The same language can sound quite different from one region to another. English spoken in the United States, the United Kingdom and India is one example; Spanish in Spain and Latin America is another.

For supported languages, you can now select a region. This helps speech recognition account for local pronunciation. It does not add an artificial accent to translated speech or change a personal voice profile. The setting only affects how the original phrase is understood.

If you are not sure which option to choose, leave the default. A specific region is most useful when Loquora regularly mishears words because of a strong regional pronunciation.

A long thought can start translating before it ends

Real conversations rarely consist of short, perfectly separated sentences. People continue a thought, pause, and add details. Waiting for the very end of a long reply can leave the other person in silence for too long.

In version 0.11.1, Loquora can translate longer speech in complete, meaningful parts. It does not simply cut the audio at fixed intervals. The app tries to find a point where part of the phrase already makes sense, sends that part for translation, and lets you continue speaking.

For ordinary segments, translation usually starts in under one second. This is not a fixed promise: connection quality, device performance, phrase length and the selected language pair all affect the result. Our translation latency page explains what we measure and what can change during a real call.

More consistent spoken output

When one long reply is translated in several parts, they need to play in the right order without talking over one another. We updated the playback queue and audio output handling so that a new part does not interrupt the previous one.

The basic call setup remains the same. The other person hears translated speech through the Loquora microphone or browser extension, while their reply can appear as translated subtitles or be read aloud. Supported paid plans can use a personal voice profile instead of a standard voice.

In-app updates on macOS

New macOS versions can now be installed from inside Loquora. When an update is ready, the app prompts you to restart and install it, so you no longer need to download a new installer for every release.

Windows users can get the current version from the download page. The same page contains the macOS app and browser extension links.

How to try the update

Install Loquora 0.11.1, choose the language you speak and the language the other person should hear, then say an ordinary sentence at your natural pace. If you are not ready to install the app, the home page now has a one-minute demo that works without an account. It lets you try recognition, translation and spoken output directly in your browser.

If a language pair or audio device behaves inconsistently, send the details through the feedback form in the app. The language, region, device model and the point where the problem appeared are especially useful.