Text to MP3: How to Turn Any Text into a Downloadable Audio File

November 3, 2026 • Tutorial
Text to MP3: How to Turn Any Text into a Downloadable Audio File

Reading text aloud on a screen is easy. Getting that speech out as a file is where people get stuck. Text to MP3 conversion sounds like a single button, but the right method depends on how much text you have, how good the voice needs to sound, and what you plan to do with the file.

This guide covers three ways to do it: using the voices already on your computer, using an online text to speech tool, and writing a short script. It also covers the problems that show up once you move past a single paragraph, like long documents, mixed languages and keeping dozens of files in order.

Why MP3 is usually the right format

You will see several audio formats when you start exporting speech. WAV and AIFF are uncompressed, so they sound clean but take up a lot of space. M4A and AAC are compressed and work well on Apple devices. MP3 is also compressed, and its main advantage is that almost everything plays it.

That matters more than people expect. A voice clip often ends up in several places: a video editor, a course platform, a podcast host, a messaging app, a phone's voicemail settings. MP3 is the format least likely to be rejected by any of them.

For speech, a bitrate of 64 to 128 kbps is plenty. Voice does not need the range that music does, and a lower bitrate keeps files small enough to email or upload quickly. If you are making an audiobook with chapters and bookmarks, M4B is the better choice, but for single clips MP3 is the safe default.

Method 1: Use your computer's built-in voices

Both macOS and Windows ship with text to speech voices. They are free and they work offline. The catch is that neither system gives you a simple "save as MP3" button, so you need a small extra step.

On a Mac

macOS includes a command-line tool called say. Open Terminal and type:

say -o greeting.aiff "Thanks for calling. Please leave a message after the tone."

That writes an AIFF file to your current folder. To read from a text file instead, use -f script.txt in place of the quoted text. You can pick a voice with -v, for example -v Samantha, and list every installed voice with say -v '?'.

The output is AIFF, not MP3. To convert it, you can open the file in an audio app and export it, or use a free command-line converter like FFmpeg:

ffmpeg -i greeting.aiff -b:a 96k greeting.mp3

On Windows

Windows does not have a one-line equivalent built in, but PowerShell can call the system speech engine and write a WAV file. Microsoft documents the method in its SpeechSynthesizer.SetOutputToWaveFile reference. A minimal version looks like this:

Add-Type -AssemblyName System.Speech
$s = New-Object System.Speech.Synthesis.SpeechSynthesizer
$s.SetOutputToWaveFile("C:\Audio\greeting.wav")
$s.Speak("Thanks for calling. Please leave a message after the tone.")
$s.Dispose()

From there, convert the WAV to MP3 the same way as on a Mac.

When this method works, and when it doesn't

Built-in voices are fine for personal notes, quick tests and internal drafts. They fall short in three places. The voices sound noticeably synthetic next to modern neural voices. The choice of languages and accents depends on what your system has installed. And the command-line steps are a barrier for anyone who doesn't live in a terminal.

Some people try a different shortcut: playing the text aloud and recording it with a screen or audio recorder. It works in a pinch, but you pick up background noise, system sounds and uneven volume, and you have to trim the start and end by hand. If the audio is going anywhere public, it is rarely worth the cleanup.

Method 2: Use an online text to speech tool

For most people, an online tool is the fastest route from text to a usable MP3. You paste the text, choose a voice, listen to a preview and download the file. There is nothing to install, and the voices are usually far more natural than the ones built into an operating system.

Here is how it works with Luvvoice:

  1. Paste or type your text into the editor.
  2. Pick a language and a voice. There are over 200 voices across 70+ languages on the free plan, and each one has a preview.
  3. Adjust the speech rate and pitch if the default delivery feels too fast or too flat.
  4. If you are signed in, insert pauses where you want a beat between sentences or sections.
  5. Generate the audio, play it back, and download it as an MP3.

A free account converts up to 3,000 characters at a time, which is roughly 450 to 500 words, and includes 10,000 free characters each month. Paid plans raise the limit to 20,000 characters per conversion, add more voices, and include commercial use rights. If the audio is going into a monetized video, an ad or a product, check the license terms of whichever tool you use before you publish.

When comparing online tools in general, a few questions are worth asking:

  • Can you download the file, or only play it? Some tools only stream audio in the browser.
  • Is there a per-conversion limit? This decides how much manual splitting you will do.
  • Is commercial use allowed on the plan you're using?
  • How many voices exist for your language? A tool with 50 English voices and one Thai voice is not much help for Thai.

Method 3: Write a short script

If you need to convert a lot of text on a schedule, such as a daily briefing or a batch of product descriptions, a script makes sense. Most text to speech providers offer an API: you send text and a voice name, and you get audio bytes back to save as a file.

A typical workflow looks like this:

  1. Store the text you want to convert in files or a spreadsheet, one row or file per clip.
  2. Loop through the rows and send each one to the API.
  3. Save each response as an MP3 with a predictable file name.
  4. Log any failures so you can retry them.

Before writing code, check the API's rate limits, how it bills (per character, per request or per minute of audio), and whether long text needs to be split before sending. Luvvoice offers API access on paid plans, with JSON responses and MP3 audio.

Handling long text without losing your place

Single sentences are easy. A 10,000-word document is where text to MP3 gets awkward.

Split on natural breaks. If a tool has a character limit, don't cut the text at an exact count. Split at the end of a paragraph or section so each file starts and ends cleanly. A chapter or heading is the best boundary.

Keep one voice and one set of settings. Write down the voice name, rate and pitch you used. Coming back three days later and choosing a slightly different setting makes the parts sound mismatched.

Listen to the joins. Play the last few seconds of one file and the first few seconds of the next. That is where awkward cuts and missing words show up.

Consider a book workflow for very long content. If you are converting a full ebook, a report or a long PDF, a tool built for that job saves a lot of splitting. The ebook to audiobook converter detects chapters automatically and exports either MP3 or M4B, so you don't have to cut and name each part yourself.

Text in more than one language

Mixed-language text causes more problems than most people expect. A voice built for English will often read a Spanish sentence with English pronunciation, and some voices will skip text in a script they do not recognize.

The cleanest approach is to split the text by language and use a native voice for each part. If you are making language learning material, for example, generate the English prompt with an English voice and the answer with a Spanish voice, then put the files side by side in your editor or playlist. It takes a few extra minutes, and the result sounds far more natural than one voice reading both languages.

Naming and organizing your MP3 files

Once you have more than a handful of files, naming becomes the thing that saves or wastes time. A simple pattern works better than a clever one:

  • Start with a number so files sort in order: 01-intro.mp3, 02-setup.mp3.
  • Use two digits (or three for large projects) so 10 doesn't sort before 2.
  • Add the language code if you produce more than one: 01-intro-en.mp3, 01-intro-es.mp3.
  • Keep the source text in a file with the same name, so you can regenerate a single clip later.

If you plan to upload the files to a podcast host or audiobook player, also fill in the title and track number in the file's metadata. Most audio editors and many free tag editors can do this in bulk.

Which method should you use?

It comes down to volume and quality:

  • A short personal clip, no rush on quality: your computer's built-in voices.
  • A clip you will publish, share or put in a video: an online text to speech tool with natural voices and direct MP3 download.
  • A full book or long document: a tool that handles chapters for you.
  • Hundreds of clips on a schedule: a script that calls a text to speech API.

Convert your first text to MP3

The quickest way to see which approach fits is to try it with something you actually need, such as a script, a lesson or a voicemail greeting. Paste a paragraph into the free online text to speech tool, pick a voice, and download the MP3. If it sounds right, you have your workflow. If it doesn't, try a different voice before changing anything else, because the voice makes the biggest difference.