This site is fictional demo content. It is not real news or affiliated with any real organization. Do not treat it as fact or professional advice.

Full article

FULL TEXT

View this issue
BriefAI

AI Real-Time Simultaneous Interpretation Protocol LinguaFlow: 108 Languages at 800-Millisecond Latency

LinguaFlow, released by Lausanne-based AI language company LinguaMind, enables real-time simultaneous interpretation across 108 languages. In European Union multilingual scenario testing, the total latency from speaker to listener receiving the translated speech was compressed to 800 milliseconds, approaching the level of human simultaneous interpreters.

LinguaFlow, released by Lausanne-based AI language company LinguaMind, is a foundational infrastructure for real-time multilingual communication. The protocol supports mutual translation across 108 languages, covering 97 percent of the world's actively used languages. Unlike traditional translation APIs, LinguaFlow is optimized for real-time spoken conversation scenarios, with end-to-end latency compressed to 800 milliseconds.

LinguaFlow adopts a streaming translation architecture that differs from the traditional batch mode of waiting until a full sentence is spoken before translating. As the speaker's voice comes in, the system begins predicting the meaning of the second half of the sentence before the full sentence is generated, and starts the translation pipeline in advance. This predictive architecture reduces total latency from the industry average of 2.4 seconds to 800 milliseconds.

The European Commission has included LinguaFlow in its Digital Service Language Fairness pilot program. In the European Parliament, the European Court of Justice, and the European Committee of the Regions, LinguaFlow will be deployed as an auxiliary translation tool. Listeners can use their phones or earbuds to choose their native language and listen in real time. LinguaFlow does not replace human simultaneous interpreters, but provides fallback translation when audiences are too large or the language is rarer.