This week we introduce the ability to generate long, multi-speaker podcasts from any number of files in Poly. Want to take a folder of legal docs and create an audio summary where each contract is introduced by a different personality? Go ahead. Want to turn your tax returns into a children’s bedtime story, complete with characters and a complex plot? Be our guest.
Poly’s Podcast mode is richer than any other major AI tool — we allow podcasts with dozens of files as context, with lengths up to 2 hours, and with multiple speakers (as many as 10 or more!). All of this comes complete with access to your file system to find the right information, and to generate things in a single simple prompt, and receive fully usable, downloadable audio files that you can take anywhere.
Unlike other AI podcast tools, you can supply long context, complex queries referencing diverse file types such as audio, video, documents, web links, and more. And you don’t need to use our app or platform to listen to the podcasts — they’re just normal audio files you can download or edit with whatever tools you want (you can also listen to them with Poly’s built in audio player of course!)
We’re extremely excited to share this feature with our growing community, as our gift for the holidays! Let us know what you think and what you’re doing with it.
<aside>
Full release notes below ↓
</aside>
Not many media publications capture the spirit of the holidays quite like The New Yorker and their tradition of printing festive covers all throughout their 100 years in operation
If you’re looking for something to listen to while preparing for your holiday celebrations this week, check out Rebecca Curtis reading her short story "The Christmas Miracle”, or why thousands of children mistakenly write letters to Satan each year. Or of course, you can generate your own podcast right in Poly…