Yesterday, WeChat officially launched a gray-scale test to integrate the DeepSeek R1 model. When entering the search bar at the top of the WeChat chat interface, an “AI Search” entry will appear on the page. After clicking it, an input box will show up, and there are two options, “Quick Q&A” and “Deep Thinking”, available for selection.

In response to this, Tencent stated that while WeChat Search – One Search enriches AI search through the use of the Hengyuan Large Model, it has recently conducted a gray-scale test to integrate DeepSeek. Users within the gray-scale test range can use the full version of the DeepSeek-R1 model for free. This integration of AI search with DeepSeek-R1 does not involve cooperation with DeepSeek and is similar to the previous launch of the DeepSeek model by Tencent Cloud. Multiple Tencent products are exploring the integration of DeepSeek.
The series of products under its banner have been integrated with DeepSeek. Tencent’s move is nothing short of swift – just last week, Tencent Cloud AI Code Assistant and Tencent Yuanbao APP also integrated with DeepSeek-R1.
It should be noted that since the industry-wide competition for large-scale models began, compared with other major players, Tencent has always been in a state of “neither too hot nor too cold”. This time, its rapid integration of DeepSeek is obviously more in line with Tencent’s cautious product strategy – the excellence of the DeepSeek model has become an industry consensus. Besides, what Tencent is more concerned about is the application implementation. As long as the application can capture users’ minds, that’s enough. Whether to use its own Hyming Large Model or those of other companies is not that important.
Major strategic initiatives of Tencent
As a “super application” with over 1.3 billion monthly active users, WeChat + DeepSeek is regarded as a “game-changer” in the industry. On one hand, after DeepSeek went public, the search scenarios that were monopolized by various independent Internet apps, or the exclusive user traffic pools, were being lost. On the other hand, from the perspective of search capabilities, this move by WeChat is both a form of defense and a significant strategic choice after the open-source route of AI large models gained the upper hand.
Why did Tencent incorporate DeepSeek – R1 into WeChat search based on the existing Hymenomai large model? What are the considerations and strategic intentions behind this move? Regarding this, a person in charge of a domestic open-source community platform told the “Sci-Tech Daily”, “The reason for WeChat to incorporate DeepSeek – R1 is multi-faceted, including technical and competitive aspects. There are also strategic conceptions from multiple dimensions such as ecology and market, which can bring benefits to Tencent in multiple fields such as users and business.”
Firstly, there is technical complementarity. DeepSeek – R1 excels in long text reasoning and complex logical deduction, which can make up for the shortcomings of the Huiyuan model in terms of real-time question answering. When dealing with “comparison and analysis type” questions, it can integrate information sources from WeChat official accounts, Video号 and the entire web to generate structured answers.
Secondly, it is about diversifying risks. Compared with its competitors who mainly promote their self-developed AI models, Tencent’s strategy of cooperating with emerging AI companies demonstrates its differentiated thinking. This approach can reduce the risk of research and development while maintaining flexibility.
Lastly, it is about leveraging the open-source feature. DeepSeek R1 is an open-source model, licensed under the MIT agreement, allowing for free commercial use, modification, and derivative development. Tencent can achieve intelligent upgrade at a lower cost by adopting this model.
“The integration of AI services has brought new commercial possibilities to Tencent. Whether it is through direct charging to users or through advertising monetization based on AI search, it can create new revenue sources for Tencent and also promote innovation in other business scenarios of WeChat, thereby driving the overall business development.” said the head of the open-source community platform.
Another industry perspective holds that Tencent has always lagged behind in technological innovation, but it likes to lay out strategic plans by leveraging its super application and huge traffic after the trend of new technologies becomes clear. From the commercial exploration of the WeChat ecosystem, this move will also reshape the ceiling of WeChat’s commercialization: B-end can output intelligent customer service, marketing copywriting generation and other capabilities through API to enterprises. C-end can also launch subscription-based AI services to activate users’ willingness to pay.