piper
Installation
SKILL.md
Piper Text-to-Speech Workflow
Convert text or markdown files to natural-sounding speech audio.
Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
input |
string | Yes | - | Path to input file (.md or .txt) or literal text |
output |
string | Yes | - | Path to output .wav file |
voice |
string | No | en_US-amy-medium |
Voice model name (see Available Voices) |
speed |
float | No | 0.67 |
Speech speed (0.5=slow, 0.67=1.5x, 1.0=normal, 1.5=fast) |
sentence_silence |
float | No | 0.3 |
Seconds of silence between sentences |
volume |
float | No | 1.0 |
Volume multiplier |
Available Voices
Models are stored in ~/piper-voices/. Each voice requires two files: .onnx and .onnx.json.