Step-Audio 2 Technisch Rapport

Step-Audio 2 Technical Report

July 22, 2025

Auteurs: Boyong Wu, Chao Yan, Chen Hu, Cheng Yi, Chengli Feng, Fei Tian, Feiyu Shen, Gang Yu, Haoyang Zhang, Jingbei Li, Mingrui Chen, Peng Liu, Wang You, Xiangyu Tony Zhang, Xingyuan Li, Xuerui Yang, Yayue Deng, Yechang Huang, Yuxin Li, Yuxin Zhang, Zhao You, Brian Li, Changyi Wan, Hanpeng Hu, Jiangjie Zhen, Siyu Chen, Song Yuan, Xuelin Zhang, Yimin Jiang, Yu Zhou, Yuxiang Yang, Bingxin Li, Buyun Ma, Changhe Song, Dongqing Pang, Guoqiang Hu, Haiyang Sun, Kang An, Na Wang, Shuli Gao, Wei Ji, Wen Li, Wen Sun, Xuan Wen, Yong Ren, Yuankai Ma, Yufan Lu, Bin Wang, Bo Li, Changxin Miao, Che Liu, Chen Xu, Dapeng Shi, Dingyuan Hu, Donghang Wu, Enle Liu, Guanzhe Huang, Gulin Yan, Han Zhang, Hao Nie, Haonan Jia, Hongyu Zhou, Jianjian Sun, Jiaoren Wu, Jie Wu, Jie Yang, Jin Yang, Junzhe Lin, Kaixiang Li, Lei Yang, Liying Shi, Li Zhou, Longlong Gu, Ming Li, Mingliang Li, Mingxiao Li, Nan Wu, Qi Han, Qinyuan Tan, Shaoliang Pang, Shengjie Fan, Siqi Liu, Tiancheng Cao, Wanying Lu, Wenqing He, Wuxun Xie, Xu Zhao, Xueqi Li, Yanbo Yu, Yang Yang, Yi Liu, Yifan Lu, Yilei Wang, Yuanhao Ding, Yuanwei Liang, Yuanwei Lu, Yuchu Luo, Yuhe Yin, Yumeng Zhan, Yuxiang Zhang, Zidong Yang, Zixin Zhang, Binxing Jiao, Daxin Jiang, Heung-Yeung Shum, Jiansheng Chen, Jing Li, Xiangyu Zhang, Yibo Zhu

cs.AI

Samenvatting

Dit artikel presenteert Step-Audio~2, een end-to-end multimodaal groot taalmodel ontworpen voor industriële audioverstaanbaarheid en spraakconversatie. Door het integreren van een latente audio-encoder en reasoning-centric reinforcement learning (RL), behaalt Step-Audio 2 veelbelovende prestaties in automatische spraakherkenning (ASR) en audioverstaanbaarheid. Om echte end-to-end spraakconversatie mogelijk te maken, incorporeert Step-Audio 2 de generatie van discrete audiotokens in taalmodeling, wat de responsiviteit op paralinguïstische informatie zoals spreekstijlen en emoties aanzienlijk verbetert. Om effectief gebruik te maken van de rijke tekstuele en akoestische kennis in real-world data, integreert Step-Audio 2 retrieval-augmented generation (RAG) en is het in staat om externe tools zoals webzoekopdrachten aan te roepen om hallucinaties te verminderen en audiozoekopdrachten om timbres te wisselen. Getraind op miljoenen uren spraak- en audiogegevens, levert Step-Audio 2 intelligentie en expressiviteit in diverse conversatiescenario's. Evaluatieresultaten tonen aan dat Step-Audio 2 state-of-the-art prestaties behaalt op verschillende audioverstaanbaarheids- en conversatiebenchmarks in vergelijking met andere open-source en commerciële oplossingen. Bezoek https://github.com/stepfun-ai/Step-Audio2 voor meer informatie.

English

This paper presents Step-Audio~2, an end-to-end multi-modal large language model designed for industry-strength audio understanding and speech conversation. By integrating a latent audio encoder and reasoning-centric reinforcement learning (RL), Step-Audio 2 achieves promising performance in automatic speech recognition (ASR) and audio understanding. To facilitate genuine end-to-end speech conversation, Step-Audio 2 incorporates the generation of discrete audio tokens into language modeling, significantly enhancing its responsiveness to paralinguistic information such as speaking styles and emotions. To effectively leverage the rich textual and acoustic knowledge in real-world data, Step-Audio 2 integrates retrieval-augmented generation (RAG) and is able to call external tools such as web search to mitigate hallucination and audio search to switch timbres. Trained on millions of hours of speech and audio data, Step-Audio 2 delivers intelligence and expressiveness across diverse conversational scenarios. Evaluation results demonstrate that Step-Audio 2 achieves state-of-the-art performance on various audio understanding and conversational benchmarks compared to other open-source and commercial solutions. Please visit https://github.com/stepfun-ai/Step-Audio2 for more information.