DuoCast turns any document into a warm two-host podcast episode. Paste text, upload a PDF or DOCX, or drop an article URL and listen on the go. Chinese and English voices are live, and episodes sound like a real conversation between two hosts.
Paste text or upload PDF and DOCX files. Convert article URLs in one click. Natural two-host conversational format. Chinese and English AI voices. Listen anywhere with shareable episode pages.
Turn long reports and papers into commute listening. Convert newsletters and articles into podcast episodes. Make study materials easier to absorb by hearing them discussed. Give written content a second life as audio.

The two-host format is the interesting choice here, and it is also where I would want to know what gets lost. What is the ratio between input and output, roughly: does a 40-page report become 10 minutes or 40, and is that length something I control? A two-host conversation is naturally lossy, it has to select and paraphrase, so the honest question for a listener is which parts get dropped. Related: does the episode page keep a link back to the source document, or a transcript alongside the audio? If I listen on a commute and want to check a number afterwards, the audio alone will not let me, and a transcript is what turns this from a nice demo into something I would use on real reports. Also worth saying on the page that Chinese and English are live and other languages are not, rather than leaving a reader to find out after uploading.
Nice direction - turning dense reports into commute listening is a real gap. Two questions from someone who tried similar flows: how do the hosts split a technical document - does one summarize while the other challenges, or is it mostly turn-taking narration? And on the sharing side, can a listener jump to a specific section of the source doc from the episode page? That would make it much easier to use episodes as an entry point back into the original material.

The two-host format is the interesting choice here, and it is also where I would want to know what gets lost. What is the ratio between input and output, roughly: does a 40-page report become 10 minutes or 40, and is that length something I control? A two-host conversation is naturally lossy, it has to select and paraphrase, so the honest question for a listener is which parts get dropped. Related: does the episode page keep a link back to the source document, or a transcript alongside the audio? If I listen on a commute and want to check a number afterwards, the audio alone will not let me, and a transcript is what turns this from a nice demo into something I would use on real reports. Also worth saying on the page that Chinese and English are live and other languages are not, rather than leaving a reader to find out after uploading.
Nice direction - turning dense reports into commute listening is a real gap. Two questions from someone who tried similar flows: how do the hosts split a technical document - does one summarize while the other challenges, or is it mostly turn-taking narration? And on the sharing side, can a listener jump to a specific section of the source doc from the episode page? That would make it much easier to use episodes as an entry point back into the original material.
Find your next favorite product or submit your own. Made by @FalakDigital.
Copyright ©2025. All Rights Reserved