杭州联汇科技股份有限公司（Om AI联汇）完成数亿元新一轮融资，由前海母基金领投。[1] 公司同时开源VLX-Seek 1.5端侧原生多模态模型，该3B参数版本在多项核心指标上超越英伟达同参数级模型。[1] 其中，模型在无人机具身场景准确率提升62.9%，目标误报率降低74.8%。[1]
本轮融资资金将用于VLX模型技术迭代、研发壁垒深化及物理AI场景商业化落地。[1] Om AI联汇已有多项产品落地应用：OttoBox AI Studio已与联想、苹果推出，[1] Homer AI可穿戴AI视觉中枢已服务近十万视障用户，月均AI调用量达千万次。[1] CEO赵天成表示，公司未来将深耕端侧原生技术创新。[1]
Om AI Lianhui (Hangzhou Lianhui Technology Co., Ltd.) has completed a new financing round valued at hundreds of millions of yuan, with Qianhai Mother Fund as the lead investor [1]. The company has open-sourced its VLX-Seek 1.5 edge-native multimodal model, which surpasses NVIDIA's same-parameter-level models across multiple core metrics [1].
The funding will be directed toward technology iterations of the VLX model, deepening research and development barriers, and commercializing edge AI applications in physical AI scenarios [1]. The company's 3-billion-parameter version of VLX-Seek 1.5 has achieved a 62.9% improvement in accuracy for unmanned aerial vehicle embodied scenarios, while reducing target false alarm rates by 74.8% [1].
Om AI Lianhui's existing products demonstrate early market traction. Its OttoBox AI Studio has been launched in partnership with Lenovo and Apple [1], while Homer AI, a wearable AI visual hub, has already served nearly 100,000 visually impaired users with an average of 10 million AI function calls per month [1]. CEO Zhao Tianchen stated that the company will continue to focus on edge-native technology innovation in the future [1].
