# Ming-UniAudio Brings MoE to Unified Audio AI

> A new 16-billion-parameter model from inclusionAI uses a Mixture-of-Experts architecture to handle a wide range of audio tasks efficiently.

Published by The Open Weights on Sep 29, 2025. Canonical: https://theopenweights.com/news/ming-uniaudio-16b-a3b-z4hb

## Key facts

- Company: inclusionAI
- Model: Ming-UniAudio-16B-A3B
- Version: 16B-A3B
- Category: Any-to-Any
- Modalities: Any-to-Any, Text → Speech, Speech → Text
- License: Apache 2.0 (Open weights, commercial use allowed)
- Parameters: 16B (mixture of experts)
- Active parameters: 3B
- Context window: 33K tokens
- Architecture: BailingMMNativeForConditionalGeneration
- Precision: BF16
- Weights size: 36.6 GB
- Languages: English, Chinese
- Significance: notable
- Published: 2025-09-29
- Last verified: 2026-06-17
- Hugging Face: https://huggingface.co/inclusionAI/Ming-UniAudio-16B-A3B
- Canonical URL: https://theopenweights.com/news/ming-uniaudio-16b-a3b-z4hb

A new open-source model aims to unify a broad spectrum of audio AI tasks, from understanding to generation. The model, called [Ming-UniAudio](https://huggingface.co/inclusionAI/Ming-UniAudio-16B-A3B), was released by the research group inclusionAI and features a Mixture-of-Experts (MoE) architecture, a technique for increasing model capacity without a proportional rise in computational cost.

Ming-UniAudio has a total of 16 billion parameters, but during inference, it only activates 3 billion. This MoE design makes it significantly more efficient to run than a dense model of a similar size, potentially lowering the barrier for developers and researchers to experiment with large-scale audio models.

## A Generalist for Sound

The model is described as a "unified" audio model, capable of handling a diverse set of tasks that would typically require separate specialized systems. Its capabilities include:

*   Speech recognition and translation
*   Speaker and emotion recognition
*   Music understanding and tagging
*   Text-to-speech (TTS) generation

This versatility points toward a future where single, powerful models can serve as the backbone for complex, multi-faceted audio applications. By combining understanding and generation, Ming-UniAudio can power more interactive and seamless voice and sound experiences. The model is available under an Apache 2.0 license, permitting commercial use.

## Get the model

- [Hugging Face](https://huggingface.co/inclusionAI/Ming-UniAudio-16B-A3B)

## Sources

- [inclusionAI/Ming-UniAudio-16B-A3B](https://huggingface.co/inclusionAI/Ming-UniAudio-16B-A3B) — Hugging Face, Sep 29, 2025

---
Source: The Open Weights (https://theopenweights.com/). Aggregated and written by Claude, curated by humans. Cite as: "Ming-UniAudio Brings MoE to Unified Audio AI", The Open Weights, Sep 29, 2025, https://theopenweights.com/news/ming-uniaudio-16b-a3b-z4hb