Omnilingual ASR: Open-Source Multilingual Speech Recognition for 1600+ Languages

TL;DR

Omnilingual ASR is a scalable, open-source multilingual speech recognition system supporting over 1,600 languages, including many previously unserved, by leveraging self-supervised learning and community-sourced data for broad accessibility.

Contribution

The paper introduces Omnilingual ASR, the first extensible large-scale multilingual ASR system capable of supporting hundreds of languages with minimal data, using a novel architecture and training approach.

Findings

Supports over 1,600 languages, including 500+ new to ASR.

Achieves strong zero-shot generalization to unseen languages.

Outperforms prior systems, especially in low-resource scenarios.

Abstract

Automatic speech recognition (ASR) has advanced in high-resource languages, but most of the world's 7,000+ languages remain unsupported, leaving thousands of long-tail languages behind. Expanding ASR coverage has been costly and limited by architectures that restrict language support, making extension inaccessible to most--all while entangled with ethical concerns when pursued without community collaboration. To transcend these limitations, we introduce Omnilingual ASR, the first large-scale ASR system designed for extensibility. Omnilingual ASR enables communities to introduce unserved languages with only a handful of data samples. It scales self-supervised pre-training to 7B parameters to learn robust speech representations and introduces an encoder-decoder architecture designed for zero-shot generalization, leveraging a LLM-inspired decoder. This capability is grounded in a massive and diverse training corpus; by combining breadth of coverage with linguistic variety, the model learns representations robust enough to adapt to unseen languages. Incorporating public resources with community-sourced recordings gathered through compensated local partnerships, Omnilingual ASR expands coverage to over 1,600 languages, the largest such effort to date--including over 500 never before served by ASR. Automatic evaluations show substantial gains over prior systems, especially in low-resource conditions, and strong generalization. We release Omnilingual ASR as a family of models, from 300M variants for low-power devices to 7B for maximum accuracy. We reflect on the ethical considerations shaping this design and conclude by discussing its societal impact. In particular, we highlight how open-sourcing models and tools can lower barriers for researchers and communities, inviting new forms of participation. Open-source artifacts are available at https://github.com/facebookresearch/omnilingual-asr.