"Translating tone and speaking pace"... DeepL Voice updated

[Financial News] Global language AI company DeepL has carried out a major update to its voice AI technology, DeepL Voice. The company said the update will support more natural communication by enabling users of Zoom, Microsoft Teams, and Google Meet to hold real-time multilingual conversations through DeepL Voice while retaining their own voices, tones, and speaking pace.
According to industry sources on the 16th, the new voice model goes a step further by preserving each speaker's unique vocal characteristics. It reflects not only "what users say" but also "how they say it."
DeepL has launched a new DeepL Voice desktop app that provides an integrated translation experience across various meeting platforms without requiring additional integrations. Through the app, voice-to-voice translation is now officially supported for online meetings, in-person conversations, and DeepL's application programming interface (API).
Real-time voice translation technology has advanced rapidly, focusing on speed and accuracy. More recently, preserving the distinctive characteristics that differentiate speakers—subtle vocal cues that affect how their spoken language is understood—has emerged as a key technical challenge. The new DeepL Voice model faithfully conveys these human elements. By retaining speech style, rhythm, pace, and intonation, it helps ensure that a real speaker's characteristics are conveyed naturally across conversations in different languages, rather than through mechanically generated voices.
Sebastian Enderlein, DeepL's chief technology officer (CTO), said, "Implementing this feature involved several technical challenges, including preserving the unique identity of each voice while ensuring both translation quality and low latency. It will not only open up new possibilities—from how companies operate and serve customers to building new multilingual products and experiences—but could also expand into a variety of use cases that we have not yet imagined."
[email protected] Jang Min-kwon Reporter