At the 2026 AI Productivity Conference, Kingsoft Office CEO Zhang Qingyuan stated that the gap in
large model capabilities is gradually narrowing, and no one can form an absolute barrier. As of June this year, Kingsoft Office's internal development has completely switched from foreign models to domestic ones, making decisions based on rational business judgment—after model capabilities become more homogenized, the model selection standard has shifted to "who is more stable and who can provide enough tokens." Zhang Qingyuan believes that model scarcity will continue to decline, with true value in the applications, workflows, and ecosystems built on these models.
IT之家：https://www.ithome.com/ 7 月 15 日消息，在今日举行的金山办公 2026 AI 生产力大会上，金山办公首席执行官章庆元就当前人工智能大模型行业的发展趋势发表了自己的看法。他指出，大模型的稀缺性正在降低，各模型之间的能力差距正逐步缩小，没有谁能形成绝对壁垒。
章庆元称：“过去三年，行业一度担心 OpenAI 会一家独大、国内模型难以追赶，但今天回头看，模型刷榜几乎每周一轮，‘某模型超越博士生’的新闻，正让人审美疲劳。”
IT之家注意到，更为关键的信号来自金山办公自身的实践。章庆元透露，公司内部开发所使用的 AI 模型已完成全面切换：去年仍大量采用国外模型，而截至今年 6 月，已全部替换为国产模型。他强调，这一决策并非出于情怀，而是基于理性的商业判断 —— 当模型能力趋向同质化，选模型的标准就从“谁更聪明”变成了“谁更稳定、谁能提供足够的 token 量”。大模型正逐步演变为类似云服务器的底层基础设施。
章庆元进一步阐释了模型与软件的本质区别。他指出，模型本身的稀缺性将持续下降，真正的价值在于基于模型构建的应用程序、工作流程及生态系统。对办公软件行业而言，AI 不是颠覆者，而是进化工具。
The readable text on this page was extracted from the public source and organized with attribution, publication time and the original link. Copyright remains with the original author and publisher.
Scan to open this article