开源证券表示，国内模型迭代节奏仍然快速，智谱综合能力全面对标Claude Opus 4.6。从参数角度来看，GLM和Kimi都具备处理处理海量资料的能力，总参数量分别为744B和1T。MiniMax总参数量为229B，略有逊色，但凭借其10B的激活参数量，模型处理速度最快可达100TPS，适用于高频Agent操作。GLM-5预训练数据为28.5T，是KimiK2.5的近两倍，知识储备量大，相较而言更为适合大型企业或全栈工程师。从OpenClaw集成水平来看，MiniMax占据先发优势，凭借首个基于OpenClaw框架搭建的云端AI一举带动国内的“龙虾热”，为国产模型接入OpenClaw提供了良好的示范作用。智谱旗下的AutoClaw则是国内首个一键安装的本地版OpenClaw应用，通过其对MCP与skill等的整合优化，成为国产Claw的有利竞品。GLM-5.1作为国产模型中率先达到在复杂目标下的持续8小时执行。
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.