开源多语言语音识别模型 Hojo-ASR-Multi-V1 发布，平均 WER 3.54
把 Qwen3-4B 接到语音识别解码端，5 种语言平均 WER 3.54，Apache 2.0 开源可本地跑，做语音转写的可以试试。
Hojo-ASR-Multi-V1 是一个开源多语音识别模型，在 5 种语言上取得 3.54 的平均 WER（词错误率）。它采用 Encoder、Adapter 加 LLM 的架构，用 Qwen3-4B 作为解码器，让转写结果同时利用声学特征和 LLM 的语言知识，从而更好地消歧发音相近的词。模型以 Apache 2.0 许可发布在 Hugging Face，可通过 hojo-asr Python 包安装，支持本地推理，输入支持 WAV 路径、音频字节流和 SCP 格式。
An open multilingual ASR model just posted 3.54 average WER across 5 languages
Hojo-ASR-Multi-V1 combines an acoustic encoder with a Qwen3-4B decoder, giving speech recognition an unusually strong language model at the final decoding stage.
- Built with a Qwen3-4B LLM decoder - Designed for multilingual conversational speech recognition - Publicly available on Hugging Face, under the Apache 2.0 license - Installable through the hojo-asr Python package - Supports local inference - Accepts WAV file paths, audio byte streams and SCP inputs
There is a bigger technical story behind the 3.54 WER number.
Traditional ASR systems rely heavily on acoustic and language modeling components designed specifically for speech, while Hojo hands the decoding stage to a Qwen3-4B LLM after the audio has been encoded and adapted. So recognition gets access to a much richer model of how words naturally fit together while it is deciding what was actually said.
What makes Hojo even more interesting is that it is reaching that level across multiple languages rather than being built around only 1 language. And it does this with an openly available model that developers can actually run and inspect themselves.
The architecture is probably the most interesting part for me.
Hojo uses an Encoder, Adapter and LLM setup with Qwen3-4B as the decoder, so the final transcript is produced using both acoustic information from the speech and the language knowledge inside the LLM.
That gives the decoder much more context for resolving words that may sound ambiguous from audio alone.
