EuroVOX: Real-Time Translation & AI Improvements

March 25, 2026 Julia Evans – Entertainment Editor Entertainment

The European Broadcasting Union (EBU) is enhancing its EuroVOX toolkit with real-time translation capabilities and improved artificial intelligence support, according to a recent announcement from EBU Technology & Innovation.

Launched in April 2024 as a Europe-wide initiative, EuroVOX aims to streamline the transcription and translation of audio and video content for broadcasters. The toolkit provides an application programming interface (API) integrating speech-to-text recognition and machine translation models. Initially available to EBU members, a public open-source release is planned.

EuroVOX is a collaborative effort between the EBU and its member public service broadcasters, designed to reduce the cost and complexity associated with multilingual media production. The system dynamically accesses different vendors, platforms, and engines, allowing users to select the most appropriate tool for specific tasks. Currently, EuroVOX supports major commercial cloud vendors and the open-source Marian translation engine, offering access to over 65 languages and dialects for transcription and nearly 20,000 language pairs for translation. The system is designed to be extensible, allowing for the integration of new vendors and technologies.

The toolkit consists of two primary components: EuroVOX Core, an API for direct integration into existing workflows for both file-based and streaming content, and EuroVOX Tool, a web-based interface for content navigation, transcription, and translation. While initially focused on file-based content, the development of capabilities to handle live content is underway, including tools to assist with live subtitling in original languages or translated versions.

Broadcasters are increasingly utilizing language tools from various vendors for diverse applications, including news monitoring, social media publishing, web transcripts, and subtitling. EuroVOX addresses the need for flexibility, recognizing that different tools may excel in specific languages or content domains. The EBU notes that the quality and performance of machine-learning-based language tools are subject to change as models are updated and refined, highlighting the importance of streamlined integration processes.

The EBU’s internal AI engine is powering the new real-time translation innovation. The EBU website notes that translations are generated with the help of AI without human review.