The Qwen team announced a major upgrade to Deep Research: in addition to generating structured research reports, it now supports publishing content as online web pages and podcast audio, catering to multi-device consumption scenarios such as "reading, viewing, and listening." The Deep Research documentation in Alibaba Cloud Model Studio explains that the agent decomposes complex problems, performs search and analysis, and produces finished products. It now supports multi-modal output, including web pages and podcasts. The Tongyi Agent product page also describes its capabilities as "web and podcast presentation."
To support the new publishing model, the relevant model capabilities are covered by Qwen3-Coder (code and automation), Qwen-Image (illustration and visualization), and Qwen3-TTS (text-to-speech). The corresponding models and features have been previously disclosed in the official blog and open source repository. Third-party reports and demonstrations indicate that the Qwen tool can quickly convert research manuscripts into online webpages for easy sharing and continuous updating.
Frequently Asked Questions
Q: What forms can Deep Research output now?
A: Research reports (long articles), web pages, and podcasts. Web pages are easy to share and update, while podcasts are convenient for listening on the commute.
Q: What are the underlying models responsible for?
A: Qwen3-Coder is responsible for code and automation processes, Qwen-Image provides image generation/editing, and Qwen3-TTS is responsible for speech synthesis in multiple languages.
Q: Where are these abilities used?
A: It can be used in the Deep Research model and Tongyi proxy capabilities of Alibaba Cloud Model Studio. Access and billing are based on the documentation.
Q: How do you understand web publishing?
A: Structure the research results into a website/page and put it online, suitable for ongoing maintenance and external reference.
Q: What are the applicable scenarios for podcast output?
A: Converting long studies into audible versions is helpful for quickly reviewing key points or learning in a screen-free setting.
