Text to MP3: save text to speech as an MP3 file
An MP3 is what most people want in the end: a file that plays on any phone, goes into a podcast, a messenger or a course. The free SuperClonica engine voices your text and saves it straight to MP3, on your own computer and without a character limit.
This guide shows how to get an MP3 from one phrase or from a whole folder of texts, which voices to use and when WAV or FLAC is the better choice.
Text to MP3 in one command
No converter and no upload site are needed: the text never leaves your computer.
- Install the engine once: python -m pip install superclonica-cpu.
- Then one command: superclonica --text "Hello! This file was made from text." --voice F1 --lang en --output hello.mp3.
- The format follows the extension: .mp3 gives an MP3 (128 kbps, 44.1 kHz, mono in the current version), .wav a 16-bit WAV, .flac and .ogg work too.
- The model downloads on the first run; after that everything works offline.
Try a voice free online first
Before you install anything, hear the voice: on the page of any voice in the catalog, type a phrase of 30 to 100 characters, choose the language and press Voice it. In a few seconds you can listen and download the recording. No account is needed.
Long texts and many files
On an ordinary 8-core desktop processor the engine is estimated at 6 to 10 times faster than real time, so an hour of audio takes several minutes.
- A whole article or chapter: --text-file chapter.txt instead of --text; long texts are split at natural pauses and joined into one file.
- A folder of texts on Linux or macOS: for f in texts/*.txt; do superclonica --text-file "$f" --voice F1 --lang en --normalize --output "${f%.txt}.mp3"; done.
- Add --normalize so every file has the same loudness, −18 LUFS.
- Name the text files the way you want the MP3s to be named.
Voices
The ten stock voices of the model, M1 to M5 and F1 to F5, are built in and free. The catalog adds narrators, conversational voices and characters, each with a free online test; free ones download without an account, and paid ones are bought once and stay yours as a JSON file you pass with --voice. Every voice speaks the 31 languages of Supertonic 3; Chinese is not supported yet.
MP3, WAV or FLAC?
MP3 is small and plays everywhere: use it for sharing, podcasts, phones and players. WAV is uncompressed: use it while you edit, so cutting and mixing do not add loss on top of loss. FLAC is lossless like WAV but smaller, good for an archive. A practical workflow is to voice and edit in WAV, then export the final file as MP3, or simply voice straight to MP3 when no editing is needed.