Text to speech on command line
This uses some new APIs in node v21.7.0 (some apis backported to v20.12.0)
Voice RSS provides a very human-sounding voices and supports 49 languages with 100 voices. See Voice RSS demo
While not nearly as good as some other TTS's out there, this one is free and good enough.
Just curl the bin/teetoes file to somewhere in your PATH & change the permissions to executable
$ curl https://raw.githubusercontent.com/Fraasi/teetoes-cli/master/bin/teetoes -o /usr/local/bin/teetoes
chmod +x /usr/local/bin/teetoesand you're good to go if you setup your api key in the config file (see below)
$ teetoes -h
teetoes v1.0.0: Text to speech on command line
Usage: teetoes [options] <text_file_path>
-l, --lang
language (default: en-us)
-v, --voice
voice (default: Linda)
-r --rate
speech rate (-10 to 10, default: 0)
-o, --output
output name for mp3 file (default: same as input file name)
-h, --help
this help
<text_file_path>
path to a text file you want to convert to audio
Note:
Options order: defaults -> env from config -> arguments
See other language & voice options at https://voicerss.org/api/demo.aspx
Issues & readme at https://github.com/fraasi/teetoes-cli
Languages have a default voice, so you can just use --lang=zh-hk to get chinese-hongkong default voice, no need to define --voice parameter.
You can specify some default configs and the mandatory api key (get it here) in the config file. Copy the config.example to ~/.config/teetoes/config and set your api key and own defaults. Empty configs will use defaults listed in help.
mkdir -p ~/.config/teetoes
curl https://raw.githubusercontent.com/fraasi/teetoes-cli/master/config.example -o ~/.config/teetoes/configJust node internals, no other dependencies.
MIT