book-to-podcast

To convert a non-fiction EPUB or PDF into a podcast RSS feed, this skill extracts the book text, generates a 35–45 minute script, renders MP3 audio via a TTS provider, and publishes updates to a private RSS feed hosted on Cloudflare R2.

Yes, the book-to-podcast pipeline supports Indian language text-to-speech by using the Sarvam provider, which enables multiple Indian languages through compatible models and specific language codes.

You can use a local TTS provider for generating audiobook-style audio by selecting the optional Kokoro provider, which runs fully locally as an alternative to paid APIs like ElevenLabs or Sarvam.

To avoid re-spending TTS credits when rerunning podcast audio generation, the skill uses resumable, per-chunk persistent audio caching so it only renders missing segments and skips already completed audio.

When turning an ebook into audio, you can generate the podcast script in either a monologue format or a two-host conversation format to structure the 35–45 minute episode.

Yes, you need Cloudflare R2 to publish your generated podcast episodes, as the skill requires R2 bucket configuration via environment variables to host and update the private RSS feed URL.