Meta Introduces SeamlessM4T: AI Model For Multilingual Speech Translation

Meta has released an AI model, SeamlessM4T, that can translate and transcribe speech in up to 100 languages.

In a post on his Instagram channel, Meta CEO Mark Zuckerberg introduced SeamlessM4T as "a new multimodal AI model that lets people who speak different languages communicate more effectively".

Zuckerberg said that the AI model can do speech-to-text, text-to-speech, speech-to-speech, text-to-text translation and speech recognition for up to 100 languages.

In order to allow researchers and developers to build on this work, SeamlessM4T is being released under a research licence, the company blog post read.

Meta plans to integrate the AI model for translation and transcription into Facebook, Instagram, WhatsApp, Messenger and Threads.

According to Meta, SeamlessM4T draws on from earlier developments in this space, such as No Language Left Behind (NLLB), a text-to-text translation model that was integrated as one of the translation providers into Wikipedia last year.