Yinmi (AudioMyst)

Yinmi can take URLs from platforms like Zhihu or WeChat public accounts, pasted text (up to 20,000 characters), or files in PDF, Markdown, TXT, and Docx formats, then automatically produce conversational podcast audio. The platform offers a selection of AI voices, allowing users to assign different voice tones to the host and guest, which makes the output feel more dynamic and natural.

It requires no audio editing experience; the whole process is guided and supports real-time preview during generation. Currently the service is free, significantly lowering the barrier for content creators to experiment with podcasting.

However, there are notable constraints: output quality can vary depending on the clarity and structure of the source material, and the AI may struggle with complex arguments or specialized terminology. The maximum input length is 20,000 characters, so very long pieces must be split. Voice options and post‐editing capabilities are basic, lacking fine‑grained controls like equalization or background music mixing. Moreover, the tool is tied to Baidu’s AI infrastructure, and the future cost model remains unclear.