NetEase LobsterAI Releases China Office Agent User Behavior Report: Daily Active Users Surge 83.1% Month-Over-Month

Deep News10-19 20:40

From October 20 to 22 along the banks of Shanghai's Huangpu River, the 2026 Sustainable Global Leaders Conference returns once again. Nobel and Turing Award winners will take the stage to discuss the future and trends of technology, and the internet-famous professors who "rescue" people from anxiety and involution — Macfarlane and Xiang Biao — will engage in a live dialogue!

Limited seats available, first come first served. Let us journey together toward sustainability! [Click to Register] Sina Tech News, evening of October 19 — Recently, NetEase (NASDAQ: NTES) LobsterAI released the "China Office Agent User Behavior Report" (September edition), focusing on September user behavior while additionally supplementing usage changes during the National Day holiday, comprehensively disclosing the latest evolutionary trends of office Agents in terms of user scale, task scale, usage habits, and model preferences.

The report shows that the monthly average daily active user scale of office Agents grew 19.8% month-over-month in June, 25.1% in July, 30.6% in August, and then soared to 83.1% month-over-month growth in September, demonstrating a clear accelerating growth trajectory.

In terms of geographic distribution, Beijing, Shanghai, Guangzhou, Shenzhen, and Hangzhou firmly held the top five positions among active user cities, while Zhengzhou, as a representative new first-tier city, successfully entered the Top 10 active user cities.

Regarding model selection, nearly 70% of users use only one model in their daily work, while heavy users tend to pair more models, using an average of 2.4 models per person. In the Top 5 list of most common model combinations, packages based on the "DeepSeek Flash series" swept all top five spots. Among them, "DeepSeek Flash series + GLM 5.3 Flash" became the most favored "golden pairing" with a 14.0% share, followed closely by "DeepSeek Flash series + V4 Pro" at 12.4%.

The report shows that thanks to the strong boost from new models, Zhipu GLM's Token share surged approximately 166.7% month-over-month in September. During the same period, several other mainstream vendors saw their shares decline to varying degrees: DeepSeek dipped slightly by 0.3%, Alibaba Qwen fell 18.2%, MiniMax dropped 66.7%, and Moonshot AI's Kimi declined 78.6%.

In the vertical multimodal generation domain, Seed series models accounted for more than half of all invocations. Specifically, in image generation tasks, Doubao Seedream dominated with a 57.2% share, far ahead of MiniMax Image (36.0%) and Alibaba Tongyi Wanxiang (6.8%); in video generation tasks, Doubao Seedance ranked first with a 54.4% share, with MiniMax Hailuo (28.1%) and Alibaba HappyHorse (17.5%) taking second and third places respectively.

The report shows that from the perspective of task distribution across industries, "general programming/IT" held an absolute 40.1% share, firmly securing its position as the largest use case for office Agents; design/film & television/creative work ranked second at 17.2%, with its share still rising, while the shares of finance, education, and internet product operations declined.

Comparing with earlier data, the scale of a single task has surged from 63,000 Tokens in March to 202,000 in September, reaching a historic peak — approximately 3.2 times that of March.

In terms of daily usage habits, WeChat has become the preferred IM channel for users to assign tasks to Agents. The report indicates that as many as 77% of users interact with Agents through WeChat, and they tend to assign tasks to Agents via IM between 7 AM and 11 AM, exhibiting the characteristic that "many people are already arranging work during their morning commute."

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.

Comments

We need your insight to fill this gap
Leave a comment