Glossary term
Glossary term
Foundations
AI task of automatically translating text from one natural language to another.
Using software (typically, a machine learning model) to convert text from one human language to another human language, for example, from English to Japanese.
DeepL uses a proprietary Transformer-based MT system that outperforms Google Translate on European language pairs in independent human-preference studies - used by 100,000+ businesses for professional document translation.
Google Translate processes over 100 billion words per day across 133 languages using neural machine translation, powering real-time camera translation in Google Lens, spoken translation in Pixel Buds, and web-page translation in Chrome.
Meta's No Language Left Behind (NLLB) model supports 200 languages including 50 African languages previously unsupported by major MT services, used by humanitarian organisations like the Red Cross for crisis communication.
Created for this library
An e-commerce company uses machine translation to localize product descriptions across more than 30 markets.
A customer support team uses machine translation to handle multilingual tickets in real time before agents respond.
A media company uses machine translation as a first pass on news content for human translator review.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License