Search
High Precision, Small Footprint, and Free for Commercial Applications: The Qwen3-ASR Speech Recognition Model Goes Fully Open-Source!
4 day ago /
Read about 0 minute
Author：小编
On January 29, Alibaba's QianWen team announced the official open-sourcing of the Qwen3-ASR series of speech recognition models. This series encompasses two distinct versions (1.7B and 0.6B parameters) along with a forced alignment model. Notably, the series boasts support for 52 languages and dialects, securing top open-source rankings across multiple authoritative benchmarks. Its performance is on par with leading closed-source models in the industry.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
