On June 2, 2026, OpenAI revealed in an official announcement a set of data: Codex's weekly active users exceeded 5 million, with non-developers accounting for about 20%, growing more than three times faster than developers. OpenAI's judgment is straightforward: "Non-developers are the real story."
At the same time, in the Chinese market, WorkBuddy's daily active users have reached 3 to 4 times that of the second-best product in the industry. This gap began to widen in March. The influx is not limited to engineers; HR, operations, and administration are becoming the main user groups.
Both products are telling the same story: enabling people who do not write code to use AI agents. However, their paths are completely different. Codex and Claude Code start from command lines and IDEs and move toward office scenarios. WorkBuddy starts from office scenarios and packages agent capabilities into tools that can be used without manuals. This difference in approach explains why non-technical users flocked to WorkBuddy first.
A Product Forced Out by Users
WorkBuddy was not a preset item on the product roadmap.
Its predecessor was CodeBuddy, an AI code assistant developed by Tencent Cloud. Normally, this team would continue to focus on developer tools. However, a deep report from Titanium Media recorded a turning point: non-technical employees from Tencent Research Institute began to spontaneously use CodeBuddy to retrieve papers and organize content. These individuals are primarily engaged in research and report writing, unrelated to coding. Team leader Wang Shengjie recognized the significance of this. His judgment is quoted by Titanium Media as: "Coding is just a process, the product is the goal."
During a weekend in mid-January 2026, Wang Shengjie and an operations colleague pulled two all-nighters to create version 0.01 of WorkBuddy. It featured a minimalist dialogue interface, pre-installed with selected skills, ready for use upon opening. There was no configuration wizard, no command line, no technical terminology. Before the public test, over 2,000 non-technical employees at Tencent were already using WorkBuddy daily; multiple media outlets, including Guangzhou Daily, confirmed this number on the day of the public test on March 9. This group’s roles were diverse, including HR, administration, operations, and sales, none of which relied on coding.
The starting point for this product line was not "we determine that office agents are a good track," but "a group of non-coders is already making do with programmer tools; let's give them something they don't have to make do with." From day one, WorkBuddy targeted non-technical users, not because market analysis said so, but because user behavior already indicated it.
Three Design Decisions
Wang Shengjie mentioned in an interview with Titanium Media: "Users don't need to understand what an agent is, what a tool is, or what prompt engineering is. They just need to know what they want."
This statement encapsulates WorkBuddy's first design decision: to replace technical concepts with natural language. In WorkBuddy, users can input "Help me organize last week's sales data into a comparative table by region," and the system will automatically break this down into data retrieval, cleaning, analysis, tabulation, and output. Users will not see terms like "agent scheduling," "tool invocation," or "context management" on the interface. In contrast, Codex and Claude Code's interaction begins with terminal commands or IDE plugins, requiring understanding of model behavior, managing token budgets, and handling execution errors—all designed for engineers, not HR.
The second decision is to pre-package scenario templates. According to reports from the Beijing News, WorkBuddy has more than 20 built-in skills packages covering data processing, invoice handling, file archiving, competitor research, content creation, public opinion analysis, and sales insights. Behind each skill is a pre-set workflow that users can use immediately without designing automation processes from scratch.
By comparison, OpenAI only introduced role plugins in Codex on June 2, 2026, with the first six covering data analysis, creative production, sales, product design, public equity investment, and investment banking. Anthropic's Claude Cowork took a different route; it does not provide pre-set templates but allows users to operate local files and applications directly on the desktop via natural language. This approach stems from an awkward discovery: Anthropic's official product page acknowledges that the internal market and data teams were bypassing the chat interface to use the developer-focused Claude Code directly.
The third decision is to make the ecosystem native rather than integrated as an add-on. Reports from Titanium Media indicate that WorkBuddy's integration with Tencent Docs is not achieved through API calls but involves "living inside." Users can directly invoke the agent to handle the current document within Tencent Docs without switching back and forth between two applications. It also supports direct remote control via WeChat, allowing tasks initiated on a computer to be tracked on a mobile phone and instructions to be supplemented before switching back to the computer to continue. For domestic companies that heavily use WeChat, Enterprise WeChat, and Tencent Docs, WorkBuddy is not "another AI tool"; it is an additional feature within their existing workflows.
Codex took a different path. OpenAI's announcement on June 2 listed 62 integrated applications from GitHub and GitLab to Salesforce, HubSpot, and Snowflake, representing a universal integration strategy aimed at the global market, with sufficient breadth but depth reliant on the quality of each third-party API. Claude Cowork focuses on operations with local file systems and desktop applications, lacking deep native integration with office suites.
The three decisions correspond to the removal of three barriers: cognitive barriers, where there's no need to understand technical concepts; scenario barriers, where users don't need to break down task flows themselves; and environmental barriers, where there is no need to leave the office software being used. The role plugins of Codex and Claude Cowork are moving in the same direction, but WorkBuddy has delivered a complete non-technical user solution ahead of schedule in the first half of 2026.
What the Growth Data Says
The changes in user structure are ultimately reflected in the numbers.
Analysys gave a clear ranking in the "2026 China Office Intelligent Agent Platform Market Research Report": as of May 2026, WorkBuddy's monthly visits on the PC side were 8.85 million, leading the second place by 2.6 times, with a month-on-month growth rate of 72.2%. Tencent’s Q1 2026 financial report confirmed this from another dimension, with DoNews citing Tencent’s statement that, based on daily active account numbers, WorkBuddy has become the most popular efficiency AI agent service in China. According to news from Investment China, WorkBuddy achieved a month-on-month growth rate of 831% in March. During the same period, Titanium Media reported that the request volume on the day of the public test far exceeded that of CodeBuddy, triggering an emergency expansion of computing resources, ultimately increasing it tenfold.
The increase in the proportion of non-technical users currently only has qualitative descriptions; "significant improvement" appeared in OmniTools' newsflash on June 16, but no specific percentage has been seen in public channels. If there were, it could be directly compared with Codex's 20% non-developer ratio. However, currently, the accuracy of the user portrait on WorkBuddy's side remains weaker than that on Codex's side.
Even so, the direction is clear. The growth curve matches the timing of the influx of non-technical users, with the growth rate after the March public test and the traffic entry effect of Tencent's ecosystem jointly pushing WorkBuddy to the top position.
Overseas Products are Turning Around
Both OpenAI and Anthropic have seen the same signal.
The data disclosed by OpenAI in the June 2 announcement itself is a signal: the growth rate of non-developers is more than three times that of developers. This is not the growth of a marginal user group; a new growth main force is forming. The six role plugins and the Codex Sites function released on the same day represent a clear "shift toward non-technology."
The story from Anthropic is even more straightforward. The opening paragraph of Claude Cowork's official product page acknowledges the product's origin: the company's internal non-technical team first "voted with their feet" by using Claude Code, which then led to Cowork. Anthropic concludes that, "most AI assistants require users to break work into individual prompts, whereas Claude Cowork accepts a result and handles the rest." Forte Labs pointed out in a comparative analysis that Claude Code requires command line installation for use and is aimed at developers. Claude Cowork provides a simplified interface but still focuses on independent desktop operations, handling local files and applications, and is not designed for scenarios like "directly editing reports in Tencent Docs."
Both products are turning around. However, tools originating from CLI or IDE need more than just functionality upgrades to transform their interaction paradigm from command line to GUI; they also need to reconstruct permission systems, governance frameworks, scenario templates, and ecosystem integration methods. This switch itself requires time. The window period that WorkBuddy secured in the first half of 2026 aligns perfectly with this transition cycle.
Two Paths, Not Which is Better
WorkBuddy chose to build agents within users' office software. WeChat, Tencent Docs, Enterprise WeChat, and QQ are tools that users are already using. The agent is not the destination; it is a capability layer embedded within the existing workflow. Codex and Claude Code chose the path of building an agent and then letting users find it. Codex exists in the terminal, in IDEs, and in the ChatGPT application. Claude Code exists only in the terminal, while Claude Cowork is on the desktop. These products have stronger core capabilities, and their models are more flexible, enabling developers to build workflows in any scenario, but non-technical users have to "reach" this agent by overcoming installation, configuration, and understanding interaction logic barriers.
Wang Shengjie's judgment is: "Users do not need to understand what an agent is; they just need to know what they want." The difference in product philosophy between this and the philosophy of Codex and Claude Code represents a fundamental divergence between the routes of "scenario encapsulation" and "underlying capabilities." The former pursues "the shortest path for users to find the agent," while the latter seeks "to maximize what the agent can help users do."
In the domestic market, the user group covered by WeChat and Enterprise WeChat is much larger than that of terminal users, making WorkBuddy's starting point naturally closer to non-technical users. In the overseas market, GitHub and terminals are the main arenas for developers, making Codex and Claude Code's starting point entirely reasonable. The influx of non-technical users is an unexpected growth that requires time to accommodate.
Pricing also reflects the same differences. WorkBuddy's personal version Lite is priced at 39 yuan/month, Standard at 99 yuan/month, Pro at 299 yuan/month, and the enterprise version SaaS flagship at 198 yuan/person/month, which both Tencent Cloud's official pricing announcement and third-party channels have confirmed. Codex requires a ChatGPT Plus subscription to start at $20/month, according to OpenAI's official pricing page. Claude Cowork also starts from a Pro plan at $20/month, as can be checked on the Claude official pricing page. WorkBuddy's starting price is lower, giving it a naturally broader reach in the domestic market.
How Long is the Leading Window?
WorkBuddy has been ahead in "scenario encapsulation" for about six months, but the pace of its competitors is not slow.
OpenAI has planned over 11 role plugin roadmaps in its June 2 announcement, and the growth rate of non-technical users is more than three times that of developers. Anthropic has positioned Claude Cowork as "for non-technical tasks," marking a strategic shift in the product. Both teams have not viewed office agents as a sideline.
WorkBuddy's enterprise version was released on June 5, 2026, which includes project features, an enterprise management backend, and 7x24 expert digital employees. The enterprise version is a crucial step in maintaining the advantage, as the expansion from individual users to organization-level procurement will require a completely different pricing model and retention logic. The accumulation of underlying model capabilities for Codex and Claude Code is objectively present, and OpenAI and Anthropic possess the world's most advanced model training capabilities; when they complete the reconstruction of the interaction layer, WorkBuddy's first-mover advantage will face more direct competition.
A 46-year-old government cloud sales professional can generate a city economic analysis report on WorkBuddy in 15 minutes, as reported by Titanium Media. An independent developer set up six digital positions, with a monthly cost of 800 yuan replacing labor costs of 50,000 to 80,000 yuan. When non-technical users do not need to learn anything to allow the agent to work independently, the gap in daily active users is not an anomaly but a natural result of differing thresholds.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
