Saved in:
Bibliographic Details
Main Authors: ai, YuanLab., :, Wu, Shawn, Wang, Sean, Li, Louie, Chen, Darcy, Wang, Allen, Luo, Jiangang, Zhao, Xudong, Shen, Joseph, Ma, Gawain, Jia, Jasper, Mao, Marcus, Wang, Claire, He, Hunter, Wang, Carol, Zhang, Zera, Wang, Jason, Shen, Chonly, Zhang, Leo, Chen, Logan, Meng, Qasim, Gong, James, Zhao, Danied, Zheng, Penn, Zhu, Owen, Yu, Tong
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.01718
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912801813954560
author ai, YuanLab.
:
Wu, Shawn
Wang, Sean
Li, Louie
Chen, Darcy
Wang, Allen
Luo, Jiangang
Zhao, Xudong
Shen, Joseph
Ma, Gawain
Jia, Jasper
Mao, Marcus
Wang, Claire
He, Hunter
Wang, Carol
Zhang, Zera
Wang, Jason
Shen, Chonly
Zhang, Leo
Chen, Logan
Meng, Qasim
Gong, James
Zhao, Danied
Zheng, Penn
Zhu, Owen
Yu, Tong
author_facet ai, YuanLab.
:
Wu, Shawn
Wang, Sean
Li, Louie
Chen, Darcy
Wang, Allen
Luo, Jiangang
Zhao, Xudong
Shen, Joseph
Ma, Gawain
Jia, Jasper
Mao, Marcus
Wang, Claire
He, Hunter
Wang, Carol
Zhang, Zera
Wang, Jason
Shen, Chonly
Zhang, Leo
Chen, Logan
Meng, Qasim
Gong, James
Zhao, Danied
Zheng, Penn
Zhu, Owen
Yu, Tong
contents We introduce Yuan3.0 Flash, an open-source Mixture-of-Experts (MoE) MultiModal Large Language Model featuring 3.7B activated parameters and 40B total parameters, specifically designed to enhance performance on enterprise-oriented tasks while maintaining competitive capabilities on general-purpose tasks. To address the overthinking phenomenon commonly observed in Large Reasoning Models (LRMs), we propose Reflection-aware Adaptive Policy Optimization (RAPO), a novel RL training algorithm that effectively regulates overthinking behaviors. In enterprise-oriented tasks such as retrieval-augmented generation (RAG), complex table understanding, and summarization, Yuan3.0 Flash consistently achieves superior performance. Moreover, it also demonstrates strong reasoning capabilities in domains such as mathematics, science, etc., attaining accuracy comparable to frontier model while requiring only approximately 1/4 to 1/2 of the average tokens. Yuan3.0 Flash has been fully open-sourced to facilitate further research and real-world deployment: https://github.com/Yuan-lab-LLM/Yuan3.0.
format Preprint
id arxiv_https___arxiv_org_abs_2601_01718
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Yuan3.0 Flash: An Open Multimodal Large Language Model for Enterprise Applications
ai, YuanLab.
:
Wu, Shawn
Wang, Sean
Li, Louie
Chen, Darcy
Wang, Allen
Luo, Jiangang
Zhao, Xudong
Shen, Joseph
Ma, Gawain
Jia, Jasper
Mao, Marcus
Wang, Claire
He, Hunter
Wang, Carol
Zhang, Zera
Wang, Jason
Shen, Chonly
Zhang, Leo
Chen, Logan
Meng, Qasim
Gong, James
Zhao, Danied
Zheng, Penn
Zhu, Owen
Yu, Tong
Artificial Intelligence
We introduce Yuan3.0 Flash, an open-source Mixture-of-Experts (MoE) MultiModal Large Language Model featuring 3.7B activated parameters and 40B total parameters, specifically designed to enhance performance on enterprise-oriented tasks while maintaining competitive capabilities on general-purpose tasks. To address the overthinking phenomenon commonly observed in Large Reasoning Models (LRMs), we propose Reflection-aware Adaptive Policy Optimization (RAPO), a novel RL training algorithm that effectively regulates overthinking behaviors. In enterprise-oriented tasks such as retrieval-augmented generation (RAG), complex table understanding, and summarization, Yuan3.0 Flash consistently achieves superior performance. Moreover, it also demonstrates strong reasoning capabilities in domains such as mathematics, science, etc., attaining accuracy comparable to frontier model while requiring only approximately 1/4 to 1/2 of the average tokens. Yuan3.0 Flash has been fully open-sourced to facilitate further research and real-world deployment: https://github.com/Yuan-lab-LLM/Yuan3.0.
title Yuan3.0 Flash: An Open Multimodal Large Language Model for Enterprise Applications
topic Artificial Intelligence
url https://arxiv.org/abs/2601.01718