This is a copy. Use Snipmill at snipmill.com.

Skip to content

Paste English text, pick a voice and press Make speech.

0 / 20,000 characters
Settings

Settings

Voice American English

1×

Checking this browser…

Runs on Snipmill's server: only the text is sent, and deleted after. How we keep it private

Good to know

How the speech is made

The page checks the text first, then Snipmill's server speaks it. The text goes through these steps:

Numbers and signs
What happens: Written out as words: "2019" as "twenty nineteen", "$26 billion", "3.2%", "9:30 a.m.", "-12°C", "St." as "Saint"
Text to voice
What happens: On Snipmill's server, a voice model speaks the text in the chosen voice
Skipped
What happens: Links are read as "link". Emoji, symbols, words in another script and runs of over 25 letters with no vowel are skipped, and a note on the bar says what. A text mostly in another script is stopped before Make; one with symbols only says "There are no words to read"
The WAV
What happens: Mono, 16-bit, made in your browser from the speech the server sends back
The MP3
What happens: 192 kbps, about 1.4 MB a minute, made from the WAV in your browser by the built-in ffmpeg encoder

Which setting does what

Voice
Choices: Anna (woman), Maya (woman), Claire (woman, soft), Grace (woman), Julia (woman), Daniel (man), Owen (man), Peter (man)
What it does: American English voices
Speed
Choices: 0.5× to 2×, 1× to start
What it does: How fast the voice speaks; the model speaks slower or faster, so the pitch stays the same

Where it runs

The voice model runs on Snipmill's server, so any browser, on a computer, phone or tablet, can make speech.

What people make

  • A 1,000-character product description: about 69 seconds of speech.
  • A blog post of 20,000 characters, read aloud for a podcast: about 23 minutes of speech, made in one go.
  • A narration for a video: make it here, then "Add Music to Video" or "Audio to Video" under "Next with this audio".

What goes to our server

Only the text goes to Snipmill's server, when you press the button that starts the job; the line under it says so. The server deletes it once it has answered and keeps no copy. The speech comes back to this page, where the WAV or MP3 is made, and it stays there until you save it. The voice samples you hear before making speech are files served by this site; playing them sends nothing.