Domestic large models see a wave of breakthroughs; the Stock Connect Internet ETF E Fund (513040) attracted 420 million in inflows over the past week

As of 10:14, the CSI HK Stock Connect Internet Index (931637) fell by 2.49%; among its constituent stocks, Mobvista rose 1.67%, Paradigm Intelligence rose 1.38%, Kingdee International rose 1.24%, ZhongAn Online rose 0.83%, and Chicego rose 0.11%, while Tencent Holdings fell 4.6%, China Literature fell 3.98%, XD Inc. fell 3.61%, NetEase Cloud Music fell 3.34%, and Kuaishou Technology fell 2.71%.

Data shows that the E Fund CSI HK Stock Connect Internet ETF (513040), which tracks the CSI HK Stock Connect Internet Index, has been favored by capital,

recording a net inflow of 38.8317 million yuan in the previous trading day. As of the previous trading day, the fund had a cumulative net inflow of 420 million yuan over the past week and 1.0158 billion yuan over the past year.

Recently, domestic large models have achieved breakthrough progress one after another, delivering impressive results. On July 17, Moonshot AI officially released the Kimi K3 model. With 2.8 trillion parameters and a 1 million token context window, it not only became the largest open-source weight model globally but also ranked third globally in the intelligence score of the independent AI evaluation agency Artificial Analysis.

Following closely on July 19, Alibaba's Tongyi Qianwen launched the Qwen3.8-Max preview version and announced that the official version would be released as open source soon. The model has a parameter scale of 2.4 trillion, and the official positioning is "possibly the most powerful model besides Fable 5".

In addition, Meituan open-sourced all weights, inference engines, and core technical documentation for its new generation foundational large model LongCat-2.0 on July 6. Notably, LongCat-2.0 is the industry's first trillion-parameter large model to complete the entire training and inference process relying entirely on domestic computing power, marking a milestone.

A research report from Open Securities points out that Kimi K3 represents a shock to the global large model industry by domestic models again. Its significance lies in marking the rapid leap of domestic models in large parameter and long context capabilities, with an open-source ecosystem approaching the level of top global closed-source models. Qwen3.8 is also worth looking forward to. The continuous breakthroughs of domestic models in technical performance, practicality, and cost-effectiveness may sustain the positive cycle of "large-scale financing + model performance leap + high growth in Token/ARR + valuation surge" for model manufacturers, further increasing the global market share of domestic models.

From a valuation perspective, the latest P/E ratio of the CSI HK Stock Connect Internet Index tracked by the E Fund CSI HK Stock Connect Internet ETF (513040) is only 20.26 times, placing it at the 7.06th percentile since the index's inception, meaning the valuation is lower than 92.94% of the range since inception.

The E Fund CSI HK Stock Connect Internet ETF (513040) closely tracks the CSI HK Stock Connect Internet Index, which selects securities of 30 listed companies involved in internet-related businesses within the scope of the HK Stock Connect as index samples. The top five weighted stocks include Alibaba, Tencent Holdings, Xiaomi Group, Meituan, and SenseTime, accounting for nearly 60% in total, with a high proportion of Hong Kong tech giants.

Related products:

E Fund CSI HK Stock Connect Internet ETF (513040), OTC feeder fund (Class A: 019313; Class C: 019314): One-click tracking of Chinese internet giants.

E Fund China Internet ETF (513050) tracks the China Internet 50 Index, reflecting the overall performance of securities of high-quality Chinese internet companies overseas, packaging core assets of overseas Chinese internet concepts with one click. Feeder funds (Class A RMB: 006327; Class C RMB: 006328; Class A USD: 006329; Class C USD: 006330).

The copyright of this article belongs to the original author/organization.

The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.