← Back to Model Beat
Open Source·Aug 18·all news from August 18, 2026

MOSS-VL Technical Report

Researchers have introduced MOSS-VL, a family of open-source vision-language models engineered for simultaneous perception and speech generation. By integrating vision processing directly into the language decoder’s architecture, the model aims to support real-time, interactive performance.

Covered by 1 source

  • AarXiv CS.AI↗Pengyu Wang, Chenkun Tan, Shaojun Zhou, Qirui Zhou, Yanxin Chen, Xingyang He, Huazheng Zeng, Jijun Cheng, Chenghao Wang, Xiaomeng Qian, Pengfei Wang, Zhan Huang, Shanqing Gao, Wei Huang, Longjun Cao, Wu Ran, Jie Liu, Changtai Zhu, Hongkai Wang, Yixian Tian, Chenghao Liu, Zhen Ye, Xinghao Wang, Botian Jiang, Guoguo Feng, Zhaoye Fei, Ruixiao Li, Mingshu Chen, Yang Gao, Qinyuan Cheng, Shimin Li, Xipeng QiuAug 18

Related stories

Open SourceChina's Zhipu Says Open-Source GLM-5.3 Outperforms Anthropic's Restricted Model in Vulnerability DetectionAug 14 · 7 sourcesOpen SourceCloudflare Cuts Astro Github Issues by 85% with AI AgentsAug 21Open SourceRepo0: Design-Driven Zero-to-All Code GenerationAug 21Open SourceLoreley: Repository-Scale Program Evolution with Quality-Diversity SearchAug 21