We’re bringing Mixhalo onto the DeepL platform to accelerate voice innovation and growth. Here’s how.

This has already been a big year for DeepL Voice. We’ve announced real-time voice-to-voice translation that means you hear in your chosen language what people are saying in theirs. Group Conversations enables multilingual, in-person conversations for groups of any size, in as many languages as needed. With DeepL Voice API, we’re enabling developers to build voice-to-text and voice-to-voice translations into customer support systems, products and platforms. And in April, Slator declared that DeepL Voice is the clear market leader on both the quality and stability of real-time translated captions.

Now, we’re building on this momentum with an addition to the DeepL platform that will further accelerate the rate of voice translation innovation. We’re bringing in the team from Mixhalo and their breakthrough real-time, ultra-low latency, audio streaming technology. It’s a platform that’s known for enabling incredible real-time audio streaming at concerts, sports events and the world’s leading global conferences, to thousands of attendees simultaneously.

Audio delivery that beats the speed of sound

To understand why we’re so excited about bringing Mixhalo onto the DeepL platform, consider this: You’re at a rock concert, watching a drummer perform a solo onstage. As he crashes his drumstick on a snare, you hear the noise streamed through your smartphone at the exact moment that you would expect to, if listening naturally. The same thing happens for every other fan in the arena. Mixhalo’s high-fidelity audio reaches audiences through their phones at the same time as soundwaves would reach them through the air. It’s audio streaming that matches the speed of sound itself, and does so at scale. In fact, Mixhalo sometimes has to subtly slow down its delivery of sound, in order to make it more natural. It’s not just real-time. It’s actually faster.

This ultra-low latency is a product of Mixhalo’s deep networking expertise and understanding of codecs, error correction and interpolation, which makes it possible to transmit audio at these types of time-bending speeds. Serving thousands of simultaneous users in live environments is hard to do, and even harder to do well. Mixhalo does both. When we work together to apply this to voice translation, things get really exciting, really quickly.

How ultra-low latency takes voice to the next level

Mixhalo builds on the speed advantage that DeepL already enjoys for voice translation. This advantage comes from superior contextual understanding that enables translations to start without waiting for sentences to finish. Further advances in our voice translation inference and flow are bringing DeepL Voice even closer to real time. Add in the Mixhalo-enhanced speed at which voice translations arrive with an audience, and the experience becomes truly seamless. 

And what can be done with the extra speed advantages that ultra-low latency brings? There’s a saying in sport that the best players always seem to have more time. The same is true in the voice translation space. 

A 500-millisecond time saving in the speed a spoken translation reaches its audience gives us inference time to do more with what that translation feels like. We can layer on personalization and voice cloning, nuanced intonation and emotion. In short, it gives us time to make that voice translation a more faithful reproduction of the person speaking. Together with Mixhalo, we’ve got more time to create the best voice translation experience available.

Helping enterprise tech businesses build faster with DeepL Voice

These are reasons enough for us to be excited about welcoming Mixhalo to DeepL. We’re equally excited, though, about what this means for how we’ll bring our solutions to more customers around the world.

For starters, Mixhalo joining the DeepL platform involves DeepL establishing our first office in San Francisco and significantly expanding our market presence in the US. 

But more so, this is a partnership whose joint value has already been proven in the market. Mixhalo Translate uses DeepL Voice API to deliver real-time translations of major global conferences and keynotes. Together, we’ve built pilots that are helping to bring real-time voice translation to customer support centers through platforms like Amazon Connect. And Mixhalo’s engineers are leveraging their insights and experience to help customers build faster with DeepL Voice API, accelerating adoption at the heart of the US tech industry.

Our Voice roadmap just got faster

Bringing the Mixhalo team and their technology to DeepL is transformative in many ways. Through hardware-free, multi-channel audio delivery, we’re able to power the next generation of real-time audio experiences at global scale. We’ll be accelerating our product roadmap to meet this opportunity, evolving DeepL Voice into a fully integrated, real-time communication layer across meetings, workshops, customer support and enterprise workflows. 

An exciting year for voice just got even better. We can’t wait to show you what we’ll build together.

Try DeepL Voice now!