DeepL launches next generation LLM that outperforms GPT-4, Google and Microsoft for translation quality
DeepL, a leading global Language Artificial Intelligence (AI) company, have announced the launch of its next generation language model, powered by a highly-specialised LLM technology built specifically for translation and editing. The latest update represents a breakthrough advancement in DeepL’s Language AI platform for businesses and sets a new industry standard for translation quality and performance.
The new solution combines three capabilities unmatched in this field:
A Specialised LLM – The DeepL solution leverages a LLM uniquely tuned for language, resulting in more human-like translations and writing for a variety of use cases with a reduced risk of hallucinations and misinformation.
Proprietary Data – Unlike general purpose models that simply train on the public internet, the DeepL model leverages over seven years of proprietary data specifically tuned for content creation and translation.
Human Model Tutoring – With a focus on quality, the DeepL model leverages thousands of hand-picked language experts specifically trained to “tutor” the model to best-in-class translation.
The latest results of this new solution significantly raise the bar for AI translation quality, with blind tests showing the DeepL translation is preferred by language experts 1.3 times more often than Google Translate, 1.7x more often than ChatGPT-4 and 2.3x more often than Microsoft.
“This is only the beginning of DeepL’s superior LLM-powered Language AI solutions for businesses,” said Jarek Kutylowski, CEO and Founder of DeepL. “We are a research-driven company to our core and are investing heavily in pushing the boundaries of translation and writing quality, efficiency, and adaptability. Our goal is to empower our customers worldwide with the industry-leading technology they need to thrive and scale globally without a language barrier standing in the way.”
“Bringing our new LLM to market is a significant milestone for our research team, however, the real proof lies in the tangible impact we can bring to our customers. When testing the new model against competitors and our classic model, professional translators noted that Google requires 2x more edits and Chat GPT requires 3x more edits to get to the same quality. For businesses who translate a large volume of documents, this truly impacts their efficiency and productivity, meaning team members can focus on tasks with high-business value,” said Stefan Mesken, VP Research, DeepL.
Since its inception in 2017, DeepL has become the go-to Language AI provider for businesses around the world, offering cutting-edge translation and writing solutions that are a critical investment for international business growth, addressing communication challenges across various domains, including internal operations and customer support. Unlike general-purpose AI systems, DeepL’s cutting-edge translation and writing solutions rely on specialised AI models specifically tuned for language, resulting in more precise translations for a variety of use cases and a reduced risk of hallucinations and misinformation. In business translation and writing, accuracy is paramount, making specialised AI models the most reliable and preferred solution for language challenges.
The introduction of DeepL’s new, next-gen model represents a significant enhancement to DeepL’s already industry-leading translation services. Blind tests conducted with leading linguists found a remarkable 1.7x improvement with the new LLM, against the old model, for combinations involving English to Japanese and Simplified Chinese. Additionally, there was a 1.4x improvement for the combination of English and German. These significant accuracy increases will save business users valuable time on manual editing, enabling them to communicate more efficiently across teams and markets.
Translations powered by the next-gen model are now available for DeepL Pro customers for translations in English, Japanese, German, and Simplified Chinese, with additional languages coming soon. The new LLM can be activated within the web translator by selecting “next-gen model”, giving users further control over the translation experience. DeepL Pro users are also protected by enterprise-grade security and compliance standards (ISO 27001 certification, GDPR/SOC 2 type 2 compliance), and no Pro translations are ever used to train its models.
