You already record every session. The transcript is the part you are missing.

Every other audience for this product has to be talked into recording their game. You already do it, on better gear than anyone else at the table, and the file is sitting right there. What you probably do not have is the text.

The problem

A show that has run fifty episodes has roughly two hundred hours of audio and no way to search a word of it. When a listener asks which episode the cursed sword turned up in, somebody scrubs. When you want to write show notes, somebody re-listens. And the listeners who need a transcript to follow along at all are simply not served.

The adoption barrier is already gone

The hard part of getting a transcript out of a tabletop session is the recording: mic placement, a room that is not a box, players who do not talk over each other. You have solved all of that already, usually with a track per person, which is the single biggest quality factor in how the transcript comes out.

That puts you in an unusual position. There is no gear to buy and no habit to build — the work is entirely on the other side of the recording, and you can test it on an episode you published two years ago.

Show notes without re-listening

The recap generator is built to produce something a table reads before the next session, which is very close to what an episode description needs: what happened, who was involved, what was left unresolved. It is a draft rather than a finished post, and you will edit it, but editing a draft built from the actual audio is a different job from writing one from memory.

Every recap is versioned, so a hand-edited episode description is never overwritten by a regeneration — see how recaps work.

A searchable archive of your own show

Once the episodes are transcribed you can ask questions across the whole run and get answers with citations back to the session they came from — which episode a name first appeared in, what a character actually promised, when a thread was dropped. For a long-running show that is the difference between a back catalogue and an archive.

The same data builds a wiki of everyone and everywhere your cast has encountered, which is the thing listeners of long shows build by hand on fan wikis. See cited campaign Q&A and the wiki.

Accessibility, said plainly

Deaf and hard-of-hearing listeners cannot access an audio-only show at all, and plenty of hearing listeners prefer to read. A transcript is the cheapest accessibility improvement an audio show can make, and it is a byproduct of what you would be doing for the show notes anyway.

One honest caveat before you promise your audience anything: what comes out is unlabelled text, not a broadcast-quality captioned script. It does not identify speakers. If you are publishing transcripts as an accessibility feature you will be doing an editing pass, and you should budget for it.

If you already have transcripts, bring them

Shows that already pay for transcription, or ran the audio through a tool themselves, can import the text directly instead of re-transcribing. Imports run the full chain — wiki, relationships, recap — and consume no transcription hours at all. Every plan includes some: two a month on the free tier, ten on Dungeon Master, unlimited on Guild Master. That is the cheapest possible way to test this against a back catalogue you have already paid to transcribe once.

This is not for you if…

You need speaker labels. Nothing here attributes a line to a cast member, and for a published transcript that is often the whole requirement. If you need a script that reads "MATT:" and "LAURA:", transcribe each of your per-speaker tracks separately somewhere else and interleave them, or budget for human transcription.

You need it fast and clean enough to publish unedited. This produces a good draft, not a finished document. A show with a copy editor will find it useful; a show that needs to publish a transcript within the hour, unread, will not.

  • You need per-speaker attribution in the output
  • You publish transcripts verbatim with no editing pass
  • You need live captions during a stream
  • Your show is scripted rather than played

Other ways people use this

  • Professional and paid GMs

    Running paid games for several tables means the worst continuity problem in the hobby — and a business reason to solve it. How the notes help.

  • West Marches and open-table campaigns

    Twenty players, rotating parties, one shared world — and nobody who was at every session. A shared record is the format’s missing piece.

  • Daggerheart tables

    Nothing in the pipeline is rules-aware, so Daggerheart transcribes and extracts like any other game. The prompts say D&D. Here is what to do about it.