On the 17th, it was learned that the AI podcast function was fully launched on the computer version of Doubao on the same day. After users upload a PDF or web link, they can generate a two-person dialogue podcast program with one click. The voice effect is highly anthropomorphic, and the dialogue is smooth and natural. It is reported that this capability is based on the voice podcast model launched by the Doubao large model team. Traditional AI podcasts are usually not colloquial enough, sound mechanical, and lack interaction. The Doubao voice podcast model relies on highly anthropomorphic voice effects, combined with spoken habits such as natural echoing in real professional podcasts, to make the dialogue effect reach the level of professional podcasts. At the same time, the model is built on a streaming model. Through an end-to-end convenient link, it only takes seconds to convert text creation to a two-person dialogue podcast, realizing a "low-cost, high-timeliness, and strong interaction" creation experience. This function has been launched on the button space and Doubao computer version. At present, the Doubao App has also started a small traffic test and will be fully launched in the near future, so that users can "listen to long texts" at any time.