Skip to content
New · the open voice benchmark is liveRead it
cantari
Convert

Text to MP3

Write the script, pick the voice, and walk away with the file every platform plays.

Painted writer's desk at night with a vintage microphone
Three steps

How does Text to MP3 work?

Step 1: Write or paste the script

Scripts up to 30,000 characters. Bracketed cues like [softly] are stage directions, and Gemini Flash performs them.

Step 2: Pick an engine and voice

Five engines, each listing its real voices, described by character rather than hype. Audition before you commit.

Step 3: Generate, play, download

Press generate, hear the take in the browser, and download the MP3 file. It is yours from that moment.

The output

Why MP3 as the output?

Turning text into an MP3 is this studio's home turf: paste a script, choose one of five engines, and the take comes back as a file you can publish anywhere. MP3 is the deliverable the rest of the world is built around; podcast hosts ingest it, every browser plays it, and every editor drops it on a timeline without complaint.

Four of the five engines, Kokoro, Grok Voice, MAI Voice 2, and Zonos, emit MP3 natively, and so do your cloned voices, so no transcoding ever touches the audio between generation and download. Gemini Flash is the deliberate exception: it produces PCM delivered as WAV, and earns the exception by acting bracketed cues instead of just reading them.

Real uses

What people turn into MP3 here

  • Podcast segmentsintros, ad reads, and narrated episodes generated from a typed script instead of a studio session.
  • Video voiceovernarration rendered as MP3 drops straight into a video editor next to the footage it describes.
  • Course audioe-learning modules re-generated per revision, so an updated paragraph never means a re-recording day.
  • Table readshearing a draft aloud on the free engine before committing to a final voice and a final cut.
The honest specifics
  • Scripts up to 30,000 characters
  • House math: about 1,000 characters is a minute of audio
  • Output: MP3 from Kokoro, Grok Voice, MAI Voice 2, Zonos, and cloned voices
  • No watermark, yours to keep
Straight answers

Text to MP3 questions, answered honestly.

How do I convert text to MP3 with an AI voice?
Open Text to Speech, paste or write the script, pick an engine and one of its voices, and generate. The take plays in the browser and downloads as a file. Kokoro generations are part of the free plan, so the first experiment costs nothing.
How much text can I convert to MP3 at once?
A single script runs up to 30,000 characters, which by the house math of about 1,000 characters to a minute comes out near half an hour of audio. Book-length work belongs in the Audiobook Studio, which takes manuscripts up to 150,000 characters.
Which engine should I pick for an MP3?
By trait: Kokoro for the fastest drafts, Grok Voice for its five distinct personas, MAI Voice 2 when you want real style and speed controls, Zonos for American and British voices. Gemini Flash is the one to know about separately; it acts bracketed cues but returns WAV rather than MP3.
Can I use the generated MP3 commercially?
Yes. Every export is yours: commercial rights included, no watermark in the audio, no hostage clauses waiting in the terms. Publish it, sell it, hand it to a client.
Keep converting

Related formats.

Text to WAVMP3 to TextAll supported formats

Want the longer read? Open the Text to Speech guide in the docs.

Your script is one generate away.

Paste it in, pick a voice, and the finished file downloads with full commercial rights. Free to start, no credit meter.

Open the studioAbout Text to Speech