We present MM-Eureka-Qwen, a multimodal reasoning model that successfully extends large-scale rule-based reinforcement learning (RL) to multimodal reasoning. Compared to the previous version of ...