Qwen Deep Research now generates webpages and podcasts from AI reports

TL;DR: Alibaba’s Qwen Team has expanded its Deep Research tool to automatically convert AI-generated research reports into interactive webpages and multi-speaker podcasts with 1-2 clicks. The proprietary feature leverages open-source models Qwen3-Coder, Qwen-Image, and Qwen3-TTS whilst providing managed hosting through Qwen Chat.

Alibaba’s Qwen Team has launched a significant expansion to its Deep Research tool, enabling users to transform comprehensive research reports into live webpages and multi-speaker podcasts within seconds. The update, announced on 21 October 2025, represents a major enhancement to the Qwen Chat interface, which competes directly with OpenAI’s ChatGPT.

Context and Background

The new functionality builds upon Qwen’s existing research capabilities by adding multi-format output options. Users can now select “Web Dev” to generate professional-grade webpages with inline graphics, or “Podcast” to create audio content featuring dynamic, multi-speaker narration. Both outputs are automatically hosted by Qwen, eliminating infrastructure configuration requirements for users.

The feature employs three open-source models—Qwen3-Coder for webpage structure, Qwen-Image for visual elements, and Qwen3-TTS for audio generation—within a proprietary managed workflow. Whilst developers with access to these open-source models could theoretically replicate similar functionality independently, Qwen provides an integrated, end-to-end experience through its hosted platform.

Technical Reality: The podcast feature offers 17 different speaker options for hosts and 7 for co-hosts, generating conversational content that discusses research findings rather than simply reading reports verbatim. Early testing suggests the audio quality remains more robotic compared to competing AI voice tools.

Looking Forward

The expansion positions Qwen Deep Research as a comprehensive content generation platform, particularly attractive to educators, content creators, and independent analysts seeking to scale research output across multiple formats. However, comparisons to Google’s NotebookLM—which recently exited beta—highlight potential differences in approach: NotebookLM focuses on organising and querying existing documents, whilst Qwen emphasises generating new research content from scratch.

The update raises questions about whether Qwen’s generalised, multi-format approach can match the depth and precision of more specialised research tools. The lack of public pricing details for Qwen3-Max and Deep Research capabilities suggests the commercial model remains under development. As AI research assistants proliferate, user priorities—whether favouring single-click publishing convenience or tight integration with existing materials—will likely determine market positioning.

Source Attribution:

Share this article