Multi-speaker podcast episodes

The podcast tool writes and voices a whole episode from a single prompt, and the quality of that episode is decided before you touch any other control, by what you put in the prompt. Give it a spine, not a topic.

A spine is a disagreement

A topic is "team software". A spine is the question the hosts argue about and where each one stands, in roughly 40 to 80 words: "Are teams over-tooled? One host thinks more software always helps; the other thinks it hides the real problem." That plans a real conversation, the way a debate needs a motion and two sides, not just a subject area. A flat list of talking points gives you hosts reciting, not talking.

Length and cast

Length goes up to 60 minutes and starts at 10. The script is planned to that length, roughly one segment per three minutes, not padded or trimmed to fit afterward. Hosts run from 1 to 4 and start at 2. Each host is voiced separately, so a two-host episode is two distinct voices trading turns, not one narrator doing both. The tool will refuse to run while two hosts share the same voice, so give each host their own.

Tone and brand

Vibe sets the conversational register. Brand weave decides how often your product comes up and defaults to subtle, so the episode does not read as one long ad. Voice engine, under advanced, sets the Pinnora Voice tier for every host. Two toggles, Jingle intro and Jingle outro, stitch your most recent jingle onto the ends, but only if a jingle already exists in the project, so make one first.

What you get back

You get one MP3 and a transcript, shown side by side. Every turn carries its start time, so clicking a line in the transcript jumps the audio to that moment. Billing follows the characters of speech synthesized, which means a longer, chattier episode costs more than a short, tight one.