A house voice
Use the same narrator every issue. Over a few months the voice becomes part of the newsletter, the way a podcast host does.
For newsletter writers
Substack Text to Speech That Gives Every Post an Audio Version
Paste a post and this Substack text to speech workflow turns it into a narrated MP3 you can attach as the audio version: one steady narrator for essays and notes, or a narrator plus a voice per character for serial fiction. You choose the voice, fix any line and keep the file.
AI audiobook samples
View all
How to add a Substack audio version in four steps
You start from the post text itself, before or after you publish; there is no link import.
- 1
Paste the post
Copy the text from your draft or the published page. Leave out subscribe buttons, share prompts and any image captions you do not want read aloud.
- 2
Pick the voice
Choose one narrator for an essay. For fiction, the speakers in quoted dialogue are detected and each gets a voice of their own.
- 3
Generate and check
Listen once through. If a name, a number or one sentence sounds off, edit that line and regenerate it alone.
- 4
Attach the MP3
Download the MP3 and add it to the post with Substack’s audio options, which at the time of writing include adding audio to a post, or use it in your podcast feed.
What essays and serials each need from Substack narration
A weekly essay and a fiction installment are different jobs. The studio handles both from the same paste box.
Quotes stay with you
In an essay, a quote with a tag like “she said” may be cast as a character. Move it back to the narrator from the line’s speaker menu, or give Other voices the narrator’s voice.
Numbers and names
Write out symbols and abbreviations you want spoken a certain way, and respell an unusual surname in the one line where it trips.
A cast for fiction
Serial fiction gets a narrator plus up to ten named characters, and the cast can be reused for every installment so voices never drift.
One post, one file
A post with no headings stays one chapter and one MP3. Lines such as “Part Two” start a new chapter, so remove them if you want a single file.
Drafts stay private
Only the lines you generate are sent to the speech model, and your writing is never used to train AI models.
Hear a narrator before you choose one for your newsletter
Play a sample, then pick a house voice for your Substack audio. You can change it before you generate the next post.
Full-cast AI audiobook
Alice's Adventures in Wonderland
Transcript
0:000:41
Playlist
Substack text to speech cost per post
At about 155 words per minute and 5.7 characters per word. Credits are charged per 1,000 characters: 6 on Lite, 9 on Standard and 13 on Premium.
| Post length | Audio length | Lite | Standard | Premium |
|---|---|---|---|---|
| 800 word note or short essay | About 5 minutes | 27 credits | 41 credits | 59 credits |
| 2,000 word essay | About 13 minutes | 68 credits | 103 credits | 148 credits |
| 4,000 word longread or installment | About 26 minutes | 137 credits | 205 credits | 296 credits |
The Starter plan’s 400 monthly credits cover about five 2,000 word posts on Lite, or nearly four on Standard. Compare plans on the pricing page, or time a draft with the words to minutes calculator.
[say with hyped, high energy like a YouTube intro]
Stop scrolling. In the next sixty seconds, I'm going to show you the trick that doubled my channel.
Setting the tone of a Substack voiceover
A note in square brackets at the start of a line shapes how it is read. Open a personal essay with [warm, conversational], slow a hard paragraph with [quietly, taking a breath], or give a punchline room with [dry]. Directions work on the Standard, Lite and Expressive models.
Inline sound tags such as [sigh] or [laugh] add a human beat inside a sentence on those same models. Use them sparingly in nonfiction; one well placed sigh in a two thousand word essay does more than ten. The AI voice emotion tags page lists what each tag does.
Before you publish Substack audio
Good habits
- Add one line to the post saying the audio version is narrated with AI voices
- Keep the same narrator every issue so regular listeners know the voice
- Use your own posts, or guest posts once the guest writer agrees
- Listen to names, numbers and quotes before you attach the file
Not a fit
- Reading other writers’ newsletters aloud without their permission
- Posts built mostly from charts, tables or embeds, which do not carry over to audio
- Putting the audio of a paid post somewhere public, which gives the paywalled text away
Why writers add Substack text to speech, and how to keep it up
Some readers are listeners. A share of any audience reads on the move, on walks, in the car, at the gym. A Substack audio version reaches them without changing how you write, and it takes minutes per post once the voice is set.
Paid perks stay paid. At the time of writing, Substack lets you limit a post to paid subscribers, and Substack audio attached to that post sits behind the same paywall. Check Substack’s current help pages for how audio, paid posts and podcast feeds work together today.
Serial fiction gets a full cast. If you publish a novel in installments, each one can sound like a produced episode with consistent character voices. The web novel to audiobook page covers serial casting, and multi voice text to speech explains how lines are split between speakers.
A year of posts can become a book. Paste your best essays with a Markdown heading such as “## The Long Walk” above each, and every essay becomes a chapter. Export an M4B with chapter markers and you have an audiobook edition of the newsletter; nonfiction audiobook narration covers voice choice for that.
Start with one post. New accounts created with Google get 100 gift credits (valid for 7 days, about 12 minutes on Standard; usage varies by text and voice tier) for a short sample, no card needed, and pasting and casting cost nothing. Try your most recent post in the AI audiobook generator and decide from there.
Substack text to speech FAQ
Does Substack have its own text to speech?
Substack has offered ways to listen to posts and to add audio to them, and its features change over time, so check its current help pages. Built-in options usually read in one voice. This Substack text to speech workflow gives you a file you control, with your choice of narrator or a full cast.
Which file should I attach to a Substack post?
The MP3. Each post with no headings becomes a single chapter, so you download one MP3 and add it to the post. M4B files are better kept for a collection of posts that you publish as an audiobook edition.
Should the post mention that the narration is AI?
It is good practice, and some platforms require disclosure, so check Substack’s current content guidelines. One line at the top or bottom of the post, such as “Audio version narrated with AI voices”, is enough for most newsletters.
Why did a quote in my essay get a different voice?
Quoted lines with a tag such as “he said” are treated as dialogue and cast to a speaker. In nonfiction you usually want the narrator to read them, so reassign those lines to the Narrator from the speaker menu, or set Other voices to the narrator’s voice.
How much time does one post’s audio take?
Pasting and casting take a minute or two, and generation takes a few minutes for a typical post. Allow time to listen through once at normal speed, which for a 2,000 word essay is about 13 minutes.
Give your next post an audio version
Paste it in and try Substack text to speech free for a short sample, no card needed.