Translated Introduces Free Adaptive MT in Matecat to Increase Linguists’ Productivity and Translation Quality

• Matecat now uses ModernMT's adaptive neural machine translation model as its default MT engine, giving everyone access to adaptive MT for their translations.
• Translators using the adaptive MT model in Matecat have increased their productivity by up to 4 times in the last seven years.
• Users with sophisticated needs can get advanced analytics, advanced glossary features, and optimization capabilities with paid ModernMT licenses.

ROME, May 7—Translated, the leading provider of AI-powered language solutions, today released an update to Matecat, its open-source computer-assisted translation (CAT) tool. The update integrates ModernMT's adaptive neural machine translation (MT) model as the default MT engine, available for free to all logged-in Matecat users. This system was recently named the leading MT software in IDC's latest MarketScape on Machine Translation and the most advanced implementation of responsive MT by CSA Research. 

Designed from the ground up to increase translator productivity, Matecat now takes the translation process to a new level by providing users with MT matches that improve in real time. The MT’s suggestions are now refined after each correction and each uploaded translation memory and adapted to the user's style and the style of the project they are working on. With adaptive MT, translators no longer have to make the same corrections, can work faster with more appropriate, contextual suggestions, and can focus more on the nuances that make the difference in overall quality.

Since 2017, Translated has been using ModernMT's adaptive MT system for all of the most challenging projects of its larger enterprise clients. Following the merger with ModernMT, Translated has integrated adaptive machine translation across its entire offering. This technology has enabled Translated to improve the quality of its translations and reduce the turnaround time for its clients. Data collected during this period shows that translators working with Translated and using adaptive MT suggestions in Matecat have registered an increase in productivity of up to 4 times while improving the quality of the translations performed. By giving translators free access to ModernMT's adaptive MT model in Matecat, Translated has been able to attract the best translators and expand its network of collaborators to over 500,000 linguists worldwide.

Commenting on the integration, CEO Marco Trombetti said: "Adaptive machine translation has been a huge step forward for Translated and for the localization industry. We believe that everyone should have the opportunity to access this sophisticated technology, which combines the accuracy of a translation model with the fluency of a large language model. We're confident that adaptive MT technology will help our entire industry produce better translations and increase productivity.”

By allowing everyone to use ModernMT, Translated will improve the adaptive MT system's support for minority languages. Thanks to the model’s structure, a small number of corrections can very quickly improve the overall quality of MT suggestions for all users.

“The full integration of ModernMT into Matecat is the realization of a vision we had back in 2010 - said Translated Chairwoman Isabelle Andreiu - Translators are the backbone of the language industry, and equipping them with the most advanced AI solution to improve their productivity is fundamental to our mission of allowing everyone to understand and be understood in their own language."

All Matecat users now have free access to ModernMT within Matecat, which comes with MyMemory, the world’s largest translation memory, preloaded. It includes context awareness for up to 100,000 words and supports 200 languages. Users with specific requirements, such as language service providers, organizations, and professionals working in critical scenarios, including legal or medical translation, can continue to use their licenses or purchase a license from ModernMT for more advanced features, including advanced glossary functionality, adaptivity in pretranslation, and optimized use of TMs based on the content being translated to maximize translation quality.

Matecat is available for free at matecat.com, with ModernMT active by default.

About Matecat

Matecat was the first web-based, open-source, AI-powered CAT tool and pioneer to integrate MT in 2015. It is the commercial version of the open-source CAT tool developed in an EU-funded research project conducted by Fondazione Bruno Kessler (FBK), Translated, the University of Edinburgh, and the University of Maine.

About ModernMT

ModernMT’s adaptive MT system was originally part of the Matecat research project, recognized as one of the most successful AI initiatives funded by the European Commission, before becoming a stand-alone research proposal. It was first developed by a consortium of Translated, FBK, the University of Edinburgh, and TAUS, funded by the EU, and later commercialized by a newly formed company that merged with Translated two years ago.

About Translated

Translated is a leading language services provider and a pioneer in the use of artificial intelligence (AI) to assist professional translators. The company was founded in 1999 by linguist Isabelle Andrieu and computer scientist Marco Trombetti to allow everyone to understand and be understood in their own language. In pursuit of its mission, Translated invented commercial context-adaptive machine translation (MT) in 2017 with ModernMT, which was named as a world-leading MT in the 2022 IDC MarketScape report. With ModernMT, Translated combines the sensitivity of over 300,000 vetted, native-speaking professional translators with the speed and capability of machines, making it possible to tackle once unimaginable localization projects. Today, Translated delivers fast, consistent, high-quality translations to more than 280,000 customers in 200 languages and over 40 subject areas.