At the DeepL Dialogues event held in Berlin, DeepL introduced its first voice translation tool, DeepL Voice, designed for real-time spoken communication. Known for its AI-powered translation solutions, the company aims to break down language barriers and boost efficiency in real-time communication with DeepL Voice, which is offered in two models designed for meetings and face-to-face interactions. Additionally, DeepL Voice enables multilingual communication in both virtual and physical settings.
DeepL Voice Announced!
DeepL CEO and founder Jarek Kutylowski stated that the tool is designed to offer high security and translation quality. Emphasizing the critical role of feedback from beta customers in the development of DeepL Voice as a reliable solution, Kutylowski shared:
“At DeepL, we are leaders in text translation, but new language-based challenges emerge in real-time voice translation. Incomplete sentences, pronunciation issues, and translation delays can lead to misunderstandings during conversation. To minimize these issues, we closely collaborated with our users during the beta program to address all these concerns.”
DeepL Voice is available to users in two distinct models:
- DeepL Voice for Meetings:
- Eliminates language barriers in virtual meetings. Participants can speak in their chosen language while others follow along with real-time translated subtitles, allowing everyone to communicate in their native language. This feature enhances comprehensibility and ensures that language does not hinder communication during meetings.
- DeepL Voice for Conversations:
- Designed for face-to-face communication on mobile devices, this model translates one party’s speech into subtitles for the other. Users can follow the translation on the same device, making communication easier with various translation modes available as needed.
Initially, DeepL Voice supports voice translation in 13 languages: German, Dutch, French, English, Spanish, Swedish, Italian, Japanese, Korean, Polish, Portuguese, Russian, and Turkish. DeepL plans to add more languages in the future.
Additionally, DeepL Translator provides subtitle translation in 33 languages. This new technology achieves a high level of accuracy thanks to AI models trained on extensive datasets that consider accent and contextual differences.
Christine Aubry, International Coordinator at Brioche Pasquier, who experienced DeepL Voice at the event, shared:
“Before DeepL Voice, it was challenging to involve our international branches in company projects. DeepL Voice for Meetings truly brought our teams together, solved communication issues across branches, and increased collaboration across the company. We’ve tried other tools that support only one language, but DeepL Voice is by far the most comprehensive solution we’ve used.”
Earlier this year, DeepL caught attention with its text translation solution, DeepL Write Pro. With its new-generation large language model (LLM), it outperformed GPT-4, Google, and Microsoft models in translation quality, and the company recently entered Forbes’ 2024 Cloud 100 list.
In May, DeepL continued its growth with a $300 million investment led by Index Ventures, reaching a $2 billion valuation. DeepL Voice is now available globally for use by companies worldwide.