Alibaba, a technology giant, has announced the addition of a new “Deep Research” feature to its AI-powered chatbot, Qwen Chat. This feature enables users to explore new topics in greater depth. The system analyzes information from over 10 sources and presents the results in the form of a comprehensive report.
To activate this feature, users first specify the topic they wish to research, and then respond to further questions posed by the bot. Subsequently, Qwen Chat compiles the gathered data and delivers it both in a conversational format and as a PDF document. For example, according to Chinese media reports, Qwen Chat has prepared an 11-page report on Japanese Edo-period woodblock prints.
However, some technical issues have been observed. In tests conducted by editors, the bot occasionally included empty or unreferenced sources. Additionally, when queried about vitamins for enhancing brain function, Qwen Chat encountered difficulties; in both Russian and English queries, the research process was left incomplete, and the bot failed to produce a full response.
Similar “Deep Research” functionalities have also begun to be implemented by other leading technology companies. OpenAI, for instance, introduced this mode on February 3, 2025. Based on the O3 model, the paid version of the feature generates multi-page reports, while a simplified free version based on the O4-mini model was launched in April.
Comparable research modes are currently available in xAI’s Grok, Google’s Gemini, the Perplexity AI search engine, and Microsoft’s Copilot platform.
