The project supports full article narration as well as spoken summaries. Full narration uses Edge TTS and needs no API key; summaries use Gemini and require your own key (though the free limit is quite generous)! You can set a maximum episode length and longer full readings split into numbered parts at natural text boundaries.
There’s a prerecorded sample on the demo page, and a simple command in the README that fetches an original sample article and generates fresh audio without an API key. The package can run locally or on a schedule in your own GitHub repository using GitHub Actions!
I wanted to turn some of my favorite newsletters into podcasts (like Seth’s Blog and A Learning A Day) and couldn’t find a free and simple solution that already existed. Let me know what you think! Demo: https://akspad.github.io/newsletter-to-podcast/