The state of machine translation for websites
Take a look at our interactive demo to learn how to translate a website in 2 minutes.
Weglot and Nimdzi Insights conducted this study to evaluate and compare a selection of leading machine translation (MT) technologies: Amazon Translate, DeepL, Google Cloud, Microsoft Translator and ModernMT. The aim was to identify what practical business impact MT can have on website translation.
The study involved a series of human reviews on five of the leading machine translation technologies across seven language pairs with a focus on usability, and evaluating performance.
Source language
en-US
Target languages
fr-FR, de-DE, es-ES, it-IT, zh-CN, ar-EG, pt-PT
Insights from the report
Engine performance highlights
Key takeaways
Based on the analysis, we concluded that comparing the baseline quality of different MT engines might become an obsolete practice.
Neural machine translation as a technology has reached a certain level of maturity and the market-leading MT providers produce decent stock quality.
- MT is fit for translating marketing content
- There is no “winner MT”, instead there are advance choices for best business impact
- To overcome perceptions, we need to understand true MT capabilities
- A localization model that relies on post-editing as a process step is old news
“The value of using MT effectively lies in having state-of-the-art engines handy. This study illustrates that website translations by contemporary NMT are highly usable and require mostly minor editing.”
Download your free copy now to scale your business successfully from the first step.
What's inside
Weglot is free — try now!
Get instant access to Weglot and see how it works for yourself. Keep using it for free, or upgrade when you’re ready to unlock premium benefits.