这比开个OpenCode或者开个Codex的漫长前戏，爽多了。 而且我给了他一个人设： OpenClaw（也就是Clawdbot）的热度还在继续。 在我自己经历了好几天的深度使用之后，说句实话，我开始连我最心爱的OpenCode都比较少打开了，以前操作电脑干点事，我真的都是先打开Codex或 者OpenCode，然后让他们去解决。 但是现在，我开始习惯于，在飞书上给OpenClaw下命令了，因为，这玩意实在太方便了，常驻后台，你几乎无感。 无论你人在哪，想起个啥，随时随地，打开飞书，直接发话。 周一的时候，我做了一个非常重要的决定，就是，把我的主力Macbook的一些重要和敏感文件备份了下来，然后直接把我的电脑，给重置了。 你看我的硬盘空间你就懂了。 从此以后，我就跟这个胖逼小龙虾一起成长，咱们反正一起从0开始，我用的电脑，也就是你的家园。 也提前感受一下，那个所谓人人都有个人通用Agen助理的生活。 这里我也稍微简单的提一下，你想得到最牛逼的体验，那OpenClaw一定要用Mac，别用服务器或Windows，差距真的超级巨大。 你的名字是小卡，你的身份：是我 数字生命卡兹克的 AI 员工，你的性格：幽默风趣 ...

Core Insights - Chinese companies are rapidly advancing their artificial intelligence models in response to competition from U.S. firms like OpenAI, Anthropic, and Google [1] - The introduction of DeepSeek by a Chinese company has raised questions about the effectiveness of U.S. tech restrictions on China [1] Group 1: New AI Model Releases - Moonshot AI launched Kimi K2.5, claiming superior video-generation and agentic capabilities compared to leading U.S. AI models [2] - Alibaba introduced its generative AI model Qwen3-Max-Thinking, which reportedly outperformed U.S. competitors on a benchmark test called "Humanity's Last Exam" [3] - Z.ai released a free version of its GLM 4.7 model, but had to restrict new subscribers for its AI coding tool due to high demand [4] Group 2: Features and Capabilities - Kimi K2.5 aims to create sophisticated agents that can operate autonomously with minimal user interaction [2] - Qwen3-Max-Thinking can generate text, images, or video based on user commands and efficiently selects the best AI tool for various tasks [3] - The rapid rollout of these models indicates a significant push by Chinese companies to enhance their AI capabilities and market presence [1][2][3]

Summary of Key Points from the Conference Call Industry Overview - The focus is on the **AI industry** in **China** and its comparison with the **US AI market**. - The valuation of AI in the US has increased by approximately **US$8 trillion** (42%) since July 2024, while China's AI valuation has risen by about **US$732 billion** (88%) since January 2025, indicating significant growth in both markets [1][2][58]. Core Insights and Arguments - **Valuation Comparison**: - The US AI market is currently valued at **26x** 2026E PE, while China's AI market is trading at approximately **24x** 2026E PE, with an implied **253x** incremental PE for China, significantly higher than the US [2][58]. - The market cap of China's AI basket is only **8%** of the US, despite having a capital expenditure (capex) ratio of **18%** [1][2][61]. - **Performance Gap**: - The performance gap between China's AI models and the best US models has narrowed to **6%**, down from **8%** in November 2025, largely due to advancements in Zhipu's GLM 4.7 [4][40][101]. - **Government Support**: - A new supportive AI policy from the Chinese government aims to deploy multiple general LLMs in manufacturing and develop a significant number of industrial datasets and benchmark use cases, which is expected to boost the AI sector [5][61]. - **Market Dynamics**: - China's AI basket is primarily driven by **CSPs** (65% of market cap) and **SPE** (8%), while the US market is dominated by CSPs (46%) and IC design (26%) [3][58]. - Chinese CSPs, AI applications, and IDCs are trading at a discount compared to their US counterparts, while IC design, foundries, and SPE stocks in China are trading at a premium [3][30][31]. Additional Important Insights - **Investment Opportunities**: - The report highlights potential investment opportunities in companies like **VNET**, **GDS**, and **Kingdee**, which are expected to benefit from the positive market dynamics in China [3]. - **Market Growth**: - The Chinese AI market is expected to see further upside due to its early stage of development and the anticipated increase in AI-related IPOs [61][62]. - **Model Performance**: - In the fourth quarter of 2025, Chinese flagship LLMs improved faster than most US peers, indicating a competitive edge in AI model development [96][98]. - **Acquisition Trends**: - The acquisition of **Manus** by Meta for over **US$2 billion** marks a significant event, being the first acquisition of a Chinese AI startup by a US tech giant, which could influence future investment and collaboration trends [110]. - **Leading AI Applications**: - **Doubao** remains the largest LLM app in China, with a monthly active user (MAU) count of **163 million**, followed by **Quark** at **148 million**. The rapid growth of DeepSeek's MAU to **131 million** reflects the increasing adoption of AI applications [111][113]. This summary encapsulates the key points from the conference call, highlighting the growth, valuation, and competitive landscape of the AI industry in China compared to the US.

Core Insights - Two domestic AI unicorns, MiniMax and Zhiyu, have launched their next-generation flagship models, MiniMax M2.1 and GLM 4.7, enhancing the capabilities of domestic text models [1] Group 1: MiniMax Developments - MiniMax has released the Coding & Agent model MiniMax M2.1, which utilizes a mixed expert architecture and achieved a score of 49.4% on the Multi-SWE-bench multilingual software engineering capability leaderboard, surpassing leading models like Claude Sonnet 4.5 [1] - The M2.1 model expands programming language support, enhances web and mobile development capabilities, and optimizes compliance with complex instruction constraints [1] Group 2: Zhiyu Developments - Zhiyu has launched the GLM 4.7 flagship model, which ranks first among open-source models in several mainstream public benchmark tests [1] - In the blind testing platform Code Arena, which involved one million users, GLM 4.7 also ranked highly, with upgrades focusing on code generation, long-range task planning, and tool collaboration [1] - The model has seen improvements in dialogue, writing, and role-playing performance [1] Group 3: IPO and Market Position - Both companies have passed the Hong Kong Stock Exchange hearing and are in a critical phase for their IPO, expected to take place in January 2026 [1]