The collection page lists Ming-3.0-audio-flash, Ming-3.0-audio-tiny, Ming-3.0-asr-tiny, Ming-3.0-tts-tiny, and Ming-3.0-tts-0.5B.
The listed tasks span text-to-speech, automatic speech recognition, and an audio model. Hugging Face displays 128B for audio-flash, 9B for audio-tiny, asr-tiny, and tts-tiny, and 1B for the repository named tts-0.5B. The collection page was updated four days earlier; individual repositories shown there were updated three or four days earlier.
The collection itself does not establish the models' licence terms, supported languages, training data, benchmark results, inference requirements, or production suitability. It also does not provide enough information to verify claims about openness beyond the repositories being listed publicly on Hugging Face. No release date, pricing, access limits, or deployment documentation is given on the collection page.