Castwright

Adoption take: promising, but it currently looks more like an early open-source project with a clear capability thesis than a broadly validated production audiobook platform. The available evidence is mostly a GitHub repo description plus a small number of stars and forks. That is attention proof, not enough usability proof for long-form stability, audio quality, or workflow maturity.

Its practical role is not a general writing copilot and not just a simple text-to-speech web app. A better analogy is a local pipeline for long-form, character-aware audiobook generation. The key claim is that an LLM assigns voices to characters and keeps them consistent across a book or series, while also supporting per-line emotion, multilingual output, and exports suited to audiobook ecosystems like M4B and Audiobookshelf.

On barriers and cost, the evidence only supports a cautious read: it is local-first and likely better suited to technical users. The GitHub text supports local operation, but there is not enough evidence to state stable hardware requirements, acceptable rendering time, official pricing, or API fees.