Skip to content
inclusionAI Models·· 9 h agoPicksAI score75

inclusionAI releases Ming-3.0-audio-flash, a 126B MoE audio-language model under MIT

inclusionAI/Ming-3.0-audio-flash

AI summary

inclusionAI released Ming-3.0-audio-flash, a native audio-language Mixture-of-Experts model with 126B total parameters and 6.5B activated per inference step, under the MIT License. It accepts audio plus text instructions and outputs text or optional synthesized speech, covering speech recognition, spoken question answering, long-audio understanding, audio reasoning and voice interaction in Chinese and English. The model reports 99.8% on Artificial Analysis's Big Bench Audio benchmark, and the repository documents Transformers, vLLM-Omni and SGLang serving setups.

Why it matters

A 126B-parameter MoE audio-language model with 6.5B activated parameters, released under MIT and reporting 99.8% on Big Bench Audio, offers a concrete reference point for open audio reasoning work.

Source:inclusionAI Models · huggingface.co