Model Releases

Tongyi Lab open-sources Fun-ASR speech recognition models with dialect and 31-language versions

Original title:QwenAudio/Fun-ASR

AI summary

Tongyi Lab has released and open-sourced the Fun-ASR end-to-end speech recognition model series. Fun-ASR-Nano-2512 has 800M parameters, is trained on tens of millions of hours of speech, and supports Chinese, English, and Japanese, 7 Chinese dialects, and 26 regional accents; Fun-ASR-MLT-Nano-2512 also has 800M parameters, is trained on hundreds of thousands of hours, and supports 31 languages.

Key Takeaways

  • Tongyi Lab has released and open-sourced the Fun-ASR end-to-end speech recognition model series.
  • Fun-ASR-Nano-2512 has 800M parameters,is trained on tens of millions of hours of speech,and…
  • Fun-ASR-MLT-Nano-2512 also has 800M parameters,is trained on hundreds of thousands of hours,and…

Quick Facts

Published
Updated
Source
通义 QwenAudio:原创语音项目
Category
Model Releases
AI score
60
Read the original Read the full article at the original source
Why it matters

原文给出两个识别模型的训练规模、语言覆盖和开源基准对比,还附部署与推理路径,读者可据此评估在中文和方言场景的可用性。

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: Tongyi Lab open-sources Fun-ASR speech recognition models with dialect and 31-language versions. December 15, 2025. https://new.nat6.net/en/news/cmtym5uhk000aro2qyie5tusu/

View all →