DeepL Voice dominates real-time translation in new global study

DeepL Voice real-time translation delivers higher quality and more stable real-time subtitles in meetings…

A new study shows that DeepL Voice delivers higher quality translations and more stable real-time subtitles in meetings in 14 language combinations, compared to Google Meet, Microsoft Teams and Zoom.

DeepL, a global AI products and research company, today presents the results of an independent comparative study evaluating the quality and stability of real-time AI translation and captioning in meetings on the most popular collaboration platforms: Google Meet, Microsoft Teams, and Zoom.

As real-time, multilingual meetings become more common for global teams, companies are increasingly relying on AI-based voice technology to facilitate collaboration, customer negotiations, and cross-border decision-making.

The study, commissioned by DeepL and conducted independently by Slator – the leading source of research and market intelligence in translation, localization, interpretation and Language AI – shows that DeepL Voice achieves the highest scores in both translation quality and subtitle stability, outperforming the built-in subtitle tools found in Google Meet, Microsoft Teams and Zoom.

In high-stakes situations—from sales calls and supplier negotiations to internal strategy meetings—even small translation errors or unstable subtitles can significantly slow down work, create confusion, or undermine important business decisions. As global business grows, AI-powered translation is no longer just a convenience, but a necessary part of the infrastructure. The strength of the infrastructure depends on two crucial factors: the quality of the AI translation and the stability of the real-time subtitles.

The most important results are:

● DeepL Voice ranks first in translation quality in human evaluations, with a quality score of 96.4/100 for DeepL Voice for Zoom Meetings and 96.3/100 for DeepL Voice for Teams, compared to 87–89 for the other platforms evaluated.

● DeepL Voice significantly reduces the number of serious errors and lowers the percentage of critical or major translation errors by an average of 76 % compared to other platforms.

● DeepL Voice has completely correct segments in 79% of cases, compared to 42% of competing tools.

● DeepL Voice has improved subtitle stability, achieving stability scores of 88.6/100 (DeepL Voice for Zoom Meetings) and 85.8/100 (DeepL Voice for Teams) in Slator's automated image-level analysis.

● DeepL Voice reduces subtitle flicker (on-screen changes/flicker) by an average of 37.6 % compared to Microsoft Teams and by an average of 54.7 % compared to Zoom.

● 96% of the professional linguists who compared the solutions prefer DeepL Voice in all blind tests.

”Language AI is becoming a central part of the infrastructure for how global companies conduct their business. In this context, accuracy and stability are not just features, but requirements,” said Jarek Kutylowski, CEO of DeepL. ”This independent comparison shows that DeepL Voice meets both requirements and maintains a level that sets a new standard for real-time communication. When professional linguists overwhelmingly prefer a particular solution, it is a clear sign of where the market is heading.”

“We looked beyond the basic accuracy results and focused on the ’human’ side of AI subtitles: readability, fluency and stability. We didn’t just want to know if the words were right at the end; we wanted to see how they behaved while a person was trying to read them. This meant we got to evaluate both the linguistic quality and how the subtitles performed on screen, and that’s where we can clearly see that DeepL sets a new standard,” says Alex Edwards, Consulting Manager at Slator.

Why stability is as important as accuracy

Slator’s report shows that even when translations are largely accurate, frequent rewriting of captions can disrupt comprehension and reduce usability in real-world encounters. To reflect how captions appear to end users, Slator measured stability by analyzing the rendered captions frame by frame on the screen, recording flicker, oscillations, and how the text changed over time.

Methodology

Slator commissioned 28 professional linguists to conduct a blind evaluation of AI-translated subtitles in 14 language combinations (7 to English, 7 from English). The report compares the default subtitle settings in Google Meet, Microsoft Teams, and Zoom with DeepL Voice for Microsoft Teams and DeepL Voice for Zoom, including the product features available to end users (such as glossaries and speech recognition). Slator states that its conclusions are independent and that the company has retained full editorial control over the methodology and results.

With our eyes set on April

These results are being announced ahead of DeepL's global spring launch on April 16, 2026, when the company will unveil extensive updates to DeepL Voice as well as expanded capabilities in automated translation and cross-platform collaboration.

About DeepL

DeepL is a global AI company building the language infrastructure that powers global business. Over 200,000 enterprise teams and millions of individuals use DeepL's Language AI platform to communicate globally, collaborate, and work across languages in real time. By combining cutting-edge AI models with enterprise-grade security and privacy, DeepL enables organizations to work seamlessly across markets and cultural boundaries. Founded in 2017 by CEO Jarek Kutylowski, DeepL now employs over 1,000 people. The company is backed by leading investors including Benchmark, IVP, and Index Ventures. Learn more at www.deepl.com.