Harness Gen AI to
Scale Your Localization
We have invested heavily in R&D to help our clients on their journey
to fully automated machine translation.
Phrase is integrating LLM-powered solutions into its AI localization platform.
Reflecting our commitment to push the boundaries of language technology innovation.
Transform translation quality and speed
We are excited to release two groundbreaking new GenAI capabilities: Auto LQA and
Phrase Next GenMT.
With the introduction of our new AI-enabled capabilities, Phrase offers immediate, automated insights into the quality of translated content at any step of the process using Auto LQA. As well as a completely new cutting-edge translation engine in Phrase Next GenMT, that combines the power of OpenAI’s latest GPT model with our Phrase NextMT technology.
Scale and automate your LQA process
Auto LQA is the ideal first step to streamlining and optimizing your human LQA process. It automatically flags errors early, saving you time and cutting costs. By allowing AI to handle the initial review, you free up your expert linguists to focus on refining and validating results, leading to faster turnarounds and up to 60% cost savings, without compromising on quality.
All new LLM-powered translation technology
Phrase Next GenMT is our revolutionary machine translation engine that fuses the power of the latest GPT model (currently GPT-4o) with our Phrase Next MT technology. Delivering unparalleled translation quality, Next GenMT also integrates your glossaries and supports tag handling, ensuring superior accuracy and consistency.
GPT models are designed to understand and process context with precision, ensuring that your translations are not just literal but convey the intended meaning and nuance perfectly. This makes Phrase Next GenMT an ideal choice for businesses that want to leverage the best LLM technology to deliver high-stakes, high-quality translations.
Note: OpenAI deletes all the processed data immediately after processing, nothing is stored with them and is also not used by them for training
More AI-powered choices with next-gen LLM integrations
Expanding our AI localization ecosystem, Phrase now integrates with DeepL next-gen and Widn.AI, giving enterprises more control over how they scale global content.
DeepL next-gen: GDPR-compliant, ultra-fast AI translation with adaptive fluency. Ideal for real-time applications like customer service.
Widn.AI: Cost-effective, scalable AI translation for high-volume customer support, ensuring fast, accurate multilingual engagement.
FREQUENTLY ASKED QUESTIONS
Here to field your questions
1. How does Auto LQA function?
Auto LQA is a feature developed incorporating the latest OpenAI technology. It can flag translation errors at the segment level using the widely accepted and industry-standard MQM quality framework. It assigns a numeric score to the translated document and determines whether the translation is a pass or a fail according to the predefined quality threshold.
2. Is there a way to compare human LQA to Auto LQA performance?
Yes, we have developed a dedicated view in Phrase Analytics which will allow you to do just that.