The "talk to your book" feature is genuinely clever - having the context window aware of where you are in the text prevents the usual "wait, you just spoiled the end" problem with most PDF chat tools.

Question: how are you handling the audiobook generation? Is it a single voice or does it switch voices for dialogue? The latter would be much harder to get right but could make fiction much more engaging.

Also curious about the business model - you mention Audible's pricing as a comparison, but they have licensing costs for actual audiobooks. Your tool seems more like a personal TTS+chat tool, which is a different value prop. Might be worth leaning into that distinction.