MMS (Massively Multilingual Speech)

Meta AI · 2023 · Speech Recognition · USA

Open in the interactive archive →

Overview

ASR/TTS for 1,100+ languages.

Details

Vendor

Meta AI

Type

Speech Recognition

Release year

2023

Jurisdiction

USA

CEO

Mark Zuckerberg

Key staffers

Yann LeCun (Chief AI Scientist), Joelle Pineau (VP AI Research), Chris Cox (Chief Product Officer)

Deployment

Cloud

Open source

Yes

Free tier

Yes

Subscription

Mostly free/open models; Meta AI features in Meta apps

Stock status

$599 (META, May 2026)

X

@MetaAI

Capabilities

DictationAudio Generation

Safety notes

Synthetic audio risks: voice cloning or impersonation misuse, copyright/style imitation concerns, and possible unsafe or biased outputs depending on deployment safeguards.

Last verified 2026-05-12 · Primary source ↗

Cite this

“MMS (Massively Multilingual Speech).” AIDb — The AI Database. Verified 2026-05-12. https://theaidatabase.com/m/mms-massively-multilingual-speech.html

More from Meta AI

AudioCraft · Chameleon · CM3Leon · Code Llama · Code Llama 13B · Code Llama 34B