Status: ongoing. Four books written, two rendered as audiobooks, book one releasing publicly.
What it is
Underfoot is a LitRPG series planned for six books. It was written in Claude Cowork, rendered into multi-voice audiobooks with about thirty distinct voices, and it is going up on my YouTube channel a chapter at a time.
I have not read any of it. That was a rule set at the start, and it has held. I am hearing the story as a listener, in the truck, the same way anyone else would.
Where it stands
Four books are drafted, 133 chapters total. Books one and two are fully rendered as audiobooks, which is 61 chapters and a little over nine hours of finished audio. Books three and four are written and waiting on voice casting before they can be rendered. Books five and six are not written yet.
Book one, The Gutter King, is releasing on YouTube on Monday, Wednesday and Friday.
What made it work
The whole thing turns on a decision made before any fiction existed. The research phase produced a set of audio-first design rules: every recurring character needs a verbal fingerprint a synthetic voice can differentiate, the voice inventory has to be budgeted rather than unlimited, system messages have to stay sparse, and chapters should be sized for a commute.
The books were written to be rendered. Almost every later technical win traces back to that. Sparse system messages kept rendering cheap. Clean speaker attribution made automated voice routing possible. Chapter lengths meant each one fit in a single batch.
What it costs
Book two came to $22.09 for 3 hours 59 minutes of finished audio. That is about 63 cents a chapter, or $5.52 per finished hour. Cover art ran about 25 cents an image.
A professionally narrated audiobook of that length usually runs $1,500 to $4,000. That is not a fair comparison, because a human narrator brings interpretation and timing that segment routing does not replicate. The gap is still two orders of magnitude.
Where I stayed in the loop
I made every casting decision from shortlists of preview clips. I approved every cover. I set the content boundaries. I never reviewed segmentation logic or tuned an audio flag.
The full write-up
The engineering detail, including the failures, is in the blog post: The six-book series I've never read.
What's next
Book three needs three voice auditions and a credit top-up before it can be rendered. Books three and four together are 72 chapters, which projects to eight or nine more hours of audio for roughly $46.