
Machine translation
Leverage the power of machine translation for accurate, fast, and reliable translations to stay ahead of your global competition.
Discover advanced MT management features within Phrase TMS and support for 30+ engines—pushing MT technology to the next level.
Create the perfect MT localization workflow with fast and cost-effective translations that don’t compromise on quality. All Phrase Translate features are embedded into Phrase TMS and integrate smoothly into your existing workflows.
Start translating immediately using our continuously growing list of fully managed engines. No developer time needed! Prefer using a different engine? Add any of the 30+ supported generic and custom engines manually.
When working with machine translation, our customers expect us to deliver consistent and cost-effective solutions. Phrase creates new business opportunities for us with its unique and innovative MT quality estimation feature, which helps us implement MT post-editing in a way that benefits both our clients and translators.
Leveraging machine translation with Phrase, we were able to achieve a 200% increase in the languages available in our help center. This has helped us stay ahead of our product team—when mobile app stores started requiring localized documentation, we were ready. With innovative features like MT Autoselect, we are confident that we are always leveraging the best possible engine for each translation—whether it's our games, marketing material, or documentation.
FREQUENTLY ASKED QUESTIONS
MTPE, or machine translation post-editing, means all segments are pre-translated. Translation memory (TM) matches are applied first, and segments that do not have TM matches are machine-translated. A human linguist then checks all pre-translated segments and makes corrections where needed. Human translation (HT) means the segments are not pre-filled, but linguists have access to TM and MT in a side panel.
Generic or “general-purpose” MT engines such as Google Translate, Microsoft Translate, and Amazon Translate are not trained on data for a particular domain or topic. Therefore, they’re ideal for general translations.
Custom MT engines are trained on data from specific domains. The result is a more accurate MT output for that kind of content, e.g. legal, or medial.
Customizable engines are trained with one customer’s translation memory data. This is the highest level of customization as the MT engine learns from human translations that were previously done for this specific customer with their tone of voice and terminology.
Resources