Meta Debuts SeamlessM4T Multilingual AI Translator

Meta has just unveiled a new tool to address cross-language communication and comprehension called SeamlessM4T, an all-in-one multimodal and multilingual AI translation model.


According to the company, this innovation enables seamless interaction through both speech and text across various languages. Meta says it has released SeamlessM4T under a research license, inviting developers and researchers to expand on this innovation.

Key features of SeamlessM4T include:

  • Speech recognition for nearly 100 languages
  • Speech-to-text translation for almost 100 input and output languages
  • Speech-to-speech translation, supporting around 100 input languages and 36 output languages (including English)
  • Text-to-text translation for nearly 100 languages
  • Text-to-speech translation, accommodating nearly 100 input languages and 35 output languages (including English)

While developing a universal language translator akin to the Babel Fish from The Hitchhiker’s Guide to the Galaxy presents challenges, SeamlessM4T marks a significant stride forward.

The model’s integrated approach mitigates errors and delays, elevating the quality and efficiency of translation. This advancement facilitates more effective cross-lingual communication, bridging language barriers.


SeamlessM4T capitalizes on insights derived from earlier projects, including the release of No Language Left Behind (NLLB), a text-to-text machine translation model for 200 languages.

Additionally, Meta demonstrated the Universal Speech Translator, the first direct speech-to-speech translation system for Hokkien, a language lacking a widely adopted writing system.

Earlier this year, Massively Multilingual Speech was unveiled, offering speech recognition, language identification, and speech synthesis across a staggering 1,100 languages.

To learn more about SeamlessM4T, visit Meta’s AI blog.

P.S. - Like our news? Support the site: become a Patreon subscriber. Or shop with our Amazon link, or buy us a coffee! We use affiliate links when possible--thanks for supporting independent media.