midashenglm-gen
MiDashengLM-Gen is an end-to-end framework that uses a pre-trained Large Language Model and audio tokenizer as the backbone, combined with per-token conditional flow matching for autoregressive, variable-length mixed-audio scene generation.
text → audio · made by Speech Team, Xiaomi MiLM Plus
Sold by
Nobody in the catalogue publishes a price for this yet.
About
midashenglm-gen — an audio model from Speech Team, Xiaomi MiLM Plus.
It takes text and returns audio. It was published in August 2026. The catalogue files it under speak. Its weights are published under APACHE-2.0, so you may run it on your own machine, or buy it from whoever serves it cheapest.
Every current figure
- Maker
- Speech Team, Xiaomi MiLM Plus
- Register
- model
- Takes
- text
- Returns
- audio
- Published
- August 2026
- Parameters
- 2.9 billion · read from its own weights
- Licence
- apache-2.0
Known as 1 name
mispeech/midashenglm-gen