DeepL Launches Aure: Real-Time Translation That Preserves Your Voice and Tone

DeepL Launches Aure: Real-Time Translation That Preserves Your Voice and Tone

2026-11-16 10:10

Introduction: Dynamic voiceprint emotion retention reduces severe translation errors by 76%. What is Aure? Aure is DeepL's speech translation model, the core driving force behind DeepL Voice.

What is Aure?

Aure is DeepL's speech translation model and the foundational engine powering DeepL Voice. Aure continuously adapts dynamically to real-time speech translation, accurately replicating speakers' intonation fluctuations, emotional transitions, stress patterns, and speaking speed—preserving the original vocal characteristics and natural expressiveness in cross-lingual communication. Aure initially supports eight languages: English, German, Spanish, French, Italian, Japanese, Portuguese, and Korean. It can be used for web conferencing, face-to-face conversations, and via Voice API, without permanently storing user voice data.

Aure

Key Features of Aure

  • Real-Time Prosody Adaptation: Continuously adjusts translation output in real time to mirror the speaker’s intonation shifts, emotional dynamics, emphasis, and variations in speaking speed.
  • Voiceprint Preservation: Utilizing advanced voice preservation technology, Aure enables cross-lingual communication using the speaker’s authentic voice identity, eliminating robotic synthetic tones.
  • Meeting Translation: Real-time translation of meeting content on platforms such as Microsoft Teams, Zoom, and Google Meet.
  • Face-to-Face Conversations: Enables real-time multilingual dialogue through DeepL Voice’s face-to-face mode, suitable for environments like reception areas and factory floors.
  • Voice API: Developers can integrate real-time speech translation into their own applications and products using the DeepL Voice API.
  • Multilingual Support: Covers English, German, Spanish, French, Italian, Japanese, Portuguese, and Korean; Chinese language support is currently not available.

Technical Principles of Aure

  • Dynamic Speech Modeling: Aure continuously analyzes a speaker’s articulation patterns during live sessions, instantly applying their original vocal traits to shape pronunciation, emotional expression, intonation, and stress when translating into other languages—injecting the spontaneity and authenticity of human speech into the translation process.
  • Differentiation from Traditional Voice Matching: Conventional voice matching techniques typically rely on a short audio sample to maintain consistency throughout a conversation. In contrast, Aure captures dynamic vocal variations and real-time responses throughout the entire interaction, representing continuous adaptive modeling rather than static one-time sample matching.
  • Speech Extraction in Noisy Environments: DeepL Voice filters out background noise, overlapping voices, and echo, isolating the primary speaker’s voice to ensure natural and lifelike translated output—even in high-noise settings.
  • Privacy Protection Mechanism: Aure dynamically matches user voices only during real-time sessions, never creating or permanently storing voice profiles. Audio data is temporarily held in memory solely during processing and is immediately deleted after session termination. Transcribed and translated content is not stored long-term and is not used for model training; all data transmissions are encrypted end-to-end.

How to Use Aure

  • Access DeepL Voice Service: Visit the official DeepL Voice website at https://www.deepl.com/zh/products/voice/aure to choose between web conference, face-to-face conversation, or Voice API versions.
  • Confirm Language Support: Aure currently supports English, German, Spanish, French, Italian, Japanese, Portuguese, and Korean. Ensure your language pair is within the supported list before use.
  • No Manual Configuration Required: Aure is pre-integrated into all DeepL Voice products. Supported languages are automatically enabled with no additional setup needed.
  • Use in Meeting Scenarios: Enable DeepL Voice in Teams, Zoom, or Google Meet to receive real-time translation of meeting audio while preserving your unique vocal characteristics.
  • Use in Face-to-Face Scenarios: Launch the DeepL Voice face-to-face mode to engage in real-time multilingual conversations—accurately extracting clear speech even in noisy environments.
  • Use in API Scenarios: Developers can call the DeepL Voice API to embed Aure’s real-time speech translation capabilities into their own apps or systems.
  • Unsupported Languages: If your language is not supported, the system will automatically revert to DeepL’s previous speech model, ensuring translation remains functional.

Core Advantages of Aure

  • Highly Realistic Dynamic Performance: Accurately reproduces intonation, emotional flow, stress, and speaking speed in real time—making it the most dynamically expressive speech AI model in DeepL’s current portfolio.
  • Beyond Voice Tone Retention: Goes beyond basic voice similarity by capturing dynamic vocal fluctuations and immediate reactions during conversation, restoring the naturalness and emotional resonance of human interaction.
  • End-to-End Real-Time Adaptation: Unlike traditional methods that rely on brief audio samples for consistency, Aure continuously analyzes and dynamically shapes vocal output throughout the full conversation.
  • Superior Translation Quality: Reduces severe translation errors by 76% compared to native tools in meetings on Teams, Zoom, and Google Meet.
  • Effective in Complex Environments: Filters background noise, competing voices, and echo to deliver natural and realistic translated speech—even in challenging acoustic conditions.

Project Repository

  • Official Website: https://www.deepl.com/zh/products/voice/aure

Application Scenarios of Aure

  • International Online Meetings: Enables real-time translation of spoken contributions in Teams, Zoom, and Google Meet, allowing multinational teams to communicate naturally using their original voices and intonations.
  • Face-to-Face Multilingual Communication: From corporate reception desks to factory workshops, enables seamless real-time dialogue between service staff and foreign clients or colleagues—even in noisy settings where clear voice extraction is critical.
  • Cross-Border Business Negotiations: Precisely conveys subtle cues such as probing, emphasis, and concessions, avoiding misinterpretations caused by flat, machine-like delivery.
  • Customer Service & Reception: Multinational customer service centers use Aure to serve clients in their native languages while retaining the agent’s authentic voice identity.
  • Application Integration: Developers can embed real-time speech translation into apps, hardware, or enterprise systems via the DeepL Voice API, enabling use cases across education, healthcare, tourism, and more.

Source: AI Tools Hub

#AI Tools

Disclaimer: Contains third-party opinions, does not constitute financial advice

Share To
X
Telegram
WeChat
QQ
Link
Recommended Reading

Step Code – The Open-Source Terminal Programming Agent by Jieyue Star

09-24
Step Code – The Open-Source Terminal Programming Agent by Jieyue Star

Hy4 Preview Lightweight Version – Tencent Hunyuan's Open-Source Hy4 Quantized Compression Model

09-02
Hy4 Preview Lightweight Version – Tencent Hunyuan's Open-Source Hy4 Quantized Compression Model

Claude Fable 5.1 – Anthropic's Latest Flagship Large Model

09-02
Claude Fable 5.1 – Anthropic's Latest Flagship Large Model

Open-source Retool is here, drag-and-drop backend with 60+ components

08-19
Open-source Retool is here, drag-and-drop backend with 60+ components

Claude Opus 5 is Here: Half the Price Approaching Fable 5

07-25
Claude Opus 5 is Here: Half the Price Approaching Fable 5

Tencent Open-Sources 260 Real-World Tasks to Test AI Workforce Capabilities

07-25
Tencent Open-Sources 260 Real-World Tasks to Test AI Workforce Capabilities

Meituan Open-Sources AI Exam Platform: 813 Minecraft Tasks

07-25
Meituan Open-Sources AI Exam Platform: 813 Minecraft Tasks