Substack Text to Speech for Fiction: Post Full-Cast Audio Episodes

Substack's own text to speech vs a produced episode

A produced episode is something you make once and attach to the post: a narrator for the prose, a distinct voice for each character, and delivery you've directed. Substack treats it as a voiceover, and a voiceover in a post overrides the automatic text to speech, so readers hear your version. It also plays where the built-in voice doesn't: on the web post, from a player in the email, and in the app.

The trade-off is effort and cost: someone has to make the file. With FullCastBook that's a few minutes per installment once the cast is set, plus credits for the characters voiced. To compare, the examples on the home page show what a cast sounds like.

Serial fiction: one installment per episode

For Substack text to speech that sounds like a show, set up one project for the serial and make each installment a chapter. Paste the chapter, check who says what, generate, and export that chapter alone as an MP3 from its chapter view. For example, an installment containing "Did you lock the shop?" Wren asked. "Twice," said Jonah. [sigh] "And then I went back to check." becomes Wren in one voice, Jonah in another, and the narrator on the tags in between.

Serialized fiction audio works best when every installment sounds like part of the same show: the same narrator, the same pacing, the same voices for the regulars. For writers whose serial also runs on Royal Road or Patreon, web novel to audiobook covers the same rhythm; for Wattpad stories, see Wattpad to audiobook.

Keep the cast across installments

Because each installment is a chapter in the same project, the cast list carries forward. A character introduced in part 3 keeps the voice you gave them in part 30. Spell names the same way every time, since detection keys on names, and give a newcomer their own voice the week they first speak. When a season ends and you start the next one, you can create a new project that copies the cast from the old one.

Export MP3 and post it as a Substack audio post

To attach the audio to the installment itself, Substack's Help Center describes a voiceover: create an article, select the headphones icon in the editor toolbar, turn on the toggle next to "This is my narration of the post", and upload your audio file. Select Insert and the voiceover appears at the top of the post. Each post can have one voiceover, and its position can't be moved.

If your publication runs a podcast, you can instead publish the installment as an episode: Substack supports MP3, WAV and AAC uploads, and every episode gets an AI-generated transcript. Since you already have the text, you can upload your own .txt transcript in its place. Either way, the MP3 exported from FullCastBook is a 192 kbps file that's ready to upload as it is.

Audio as a paid-subscriber perk

Substack fiction audio makes a natural reward for paying readers. Substack describes two ways to do it. Keep the text free and paywall only the voiceover: add a paywall at the end of the post, set the post to Paid subscribers only and keep the free preview selected, and free subscribers get the text while the voiceover plays for paying ones. Or, for a podcast episode, upload a separate free preview file for free listeners.

Another option is to make the produced audio the paid tier and leave Substack's automatic text to speech, where your publication has it, for everyone else. Whichever you choose, say what paid subscribers get in your welcome email and about page so the perk is visible.

Rights and AI disclosure

Voice fiction you wrote or have the rights to. If you co-write or adapt someone else's work, get their agreement before you publish audio of it. The files you export are yours to use under our terms, including in a paid publication.

On disclosure: the Substack Help Center pages we read in October 2026 don't set a rule specific to AI narration. They do let you add a "How I make this" statement in your publication's settings, and the voiceover toggle is worded "This is my narration of the post". The honest, simple choice is a one-line note at the top of each post, such as "Audio narrated with AI voices." Free credits on sign-up cover a sample installment; regular episodes need a plan, see pricing. For short pieces rather than a serial, see text to audiobook or AI voice for storytelling.