AI

Audio narration turns a chapter into a narrated audio version you can play in your browser or download as an MP3. My Spicy Vanilla reads the narration with a main narrator voice and, if you want, gives every character their own voice with emotions that match the dialogue.

You start it from the chapter page with the Listen to this story button. Each generation uses one audio credit, and the voices are chosen once in your settings, so the first step is setting them up.

Step 1: choose the narrator voice

The narrator reads every line that is not spoken by a character. Set it up once and it applies to all your stories.

  1. Go to Content Preferences → Audio Generation.
  2. In the Main Narration section, pick a language in Select language, then a voice in Select voice. Voices are grouped by gender and there are hundreds to choose from.
  3. Press the Preview voice button next to the dropdown to hear a short sample before you decide.
  4. Set the Voice Speed, from Very Slow (0.5x) to Ultra Fast (2x). Normal (1x) is a safe default.
  5. Click Update.

Important: the voice must be in the same language as the story. If you write in French, pick a French voice; audio can only be generated when every voice matches the story's language.

Some voices carry a warning symbol. A voice marked as unstable may not always work, and a voice marked as deprecated will stop working soon, so prefer voices without a warning.

Step 2 (optional): give each character their own voice

By default the narrator reads everything, including dialogue. With character voices on, every character speaks with their own voice and the AI adds emotion to each spoken line automatically, based on how the line is written — a line ending in "she whispered sadly" is read sadly.

  1. On the same Audio Generation page, turn on Enable Character Voices in the Character Voices section and click Update.
  2. Open Characters and edit a character. A new Voice (for audio stories) field appears with the same language and voice dropdowns, the Preview voice button and a Voice Speed setting.
  3. Pick a voice that fits the character and save. Repeat for every character in the story.

Voices are saved on the character, not on the story, so a character keeps the same voice in every chapter and every story they appear in.

Good to know: when character voices are on, every character selected in the story needs a voice, otherwise the generation is blocked with a reminder to assign the missing ones. Character voices work best for chapters of 2,000–3,000 words; for longer chapters a single narrator is smoother and faster to generate, so consider turning the setting off for those.

Step 3: generate the audio

  1. Open the chapter from My Stories and click Listen to this story at the top of the page.
  2. The dialog shows how many audio credits you have and which voices will be used. Each voice is a link, so you can jump straight to the settings page if something needs changing.
  3. Click Generate Audio Version. You are taken to a progress page that says Generating Story Audio...; it usually takes a few minutes, so leave the page open and don't refresh it.
  4. When the audio is ready you are sent back to the chapter, where an audio player now sits at the top of the page.

Only one audio generation runs at a time per account. If you start a second one, the dialog asks you to wait for the first to finish.

Listen, download and regenerate

  • Listen: use the player at the top of the chapter. Only you, as the author, can see it.
  • Download: open the Download menu on the chapter page and choose Audio version to save the MP3 file. Before you generate the audio, the same entry shows Generate audio first.
  • Regenerate: once a chapter has audio, the Listen to this story button becomes a small round button next to the player. It opens the Regenerate Audio Version dialog. Regenerating uses one more credit, and the current version stays available until the new one is ready.

Important: editing the chapter text does not update the audio. If you change the text after generating, regenerate the audio to hear the new version.

How audio credits work

  • One credit narrates one full chapter, whatever its length.
  • If you have no credits, the Listen to this story dialog plays a real example of how a narrated story sounds and lets you buy a package of 5, 10 or 20 credits. Credits never expire.
  • Premium subscribers get 2x the credits for the same price on every package. The Premium day pass also includes 1 credit and the monthly plan includes 2.
  • A credit is only spent on a successful generation. If the generation fails, the credit is returned to your account.

Requirements and limits

  • Chapter length: audio is available for chapters under 6,000 words. For longer chapters the dialog offers an Edit Story button so you can shorten the text first.
  • Paragraph length: every paragraph must be under 5,000 characters. If one is longer, add line breaks to split it.
  • Language: the story's language must be supported for audio, and every voice must be in that language. Otherwise the dialog tells you which voices to update. Your credits stay safe in the meantime.
  • Text in brackets: anything written inside parentheses or square brackets is skipped when the audio is read, so keep author notes out of the chapter text.

If a generation is blocked, the dialog always tells you why and offers a button that takes you to the right place. For more fixes, see the troubleshooting guide.

Final tip

For your first narration, keep it simple: pick one narrator voice you like, leave character voices off, and generate a chapter of 2,000–3,000 words. Once you know how it sounds, turn on character voices and cast your characters. Well-structured dialogue with clear speech tags gives the best result — the story prompts tutorial shows how to ask the AI for it.

Love Reddit? So do we!

Join our new subreddit community to share tips, tricks, and experiences. Let's make the most out of My Spicy Vanilla together!