espeak-ng

Speak text

$ espeak-ng "[Hello world]"

$ espeak-ng -v [en-gb] "[Hello]"

$ espeak-ng -f [document.txt]

$ espeak-ng -w [output.wav] "[Hello]"

$ espeak-ng -s [175] "[Hello]"

$ espeak-ng --voices

$ espeak-ng -x "[h@l'oU]"

espeak-ng [options] [words]

espeak-ng (eSpeak New Generation) is a fork of eSpeak with active development, additional features, and improved voice quality. It's a formant-based speech synthesizer supporting 100+ languages.
The tool provides text-to-speech capabilities for accessibility, voice assistants, and applications. It includes improvements in pronunciation rules, language support, and phoneme handling over the original eSpeak.
espeak-ng is the default TTS engine in many Linux distributions and speech synthesis frameworks.

WORDS

Text to speak.

Select voice/language.

Read from file.

Write to WAV file.

Words per minute.

Pitch adjustment.

Input is phonemes.

List voices.

Display help information.

Still sounds robotic (formant synthesis). Some advanced SSML features unsupported. Voice quality varies by language. Output format options limited.

espeak-ng was forked from eSpeak by Reece H. Dunn to continue development after the original project became inactive. It's now the actively maintained version used in most Linux distributions.

espeak(1), festival(1), mimic(1)