Piper-Reader

Privacy-focused local text-to-speech reader using Piper TTS and Haskell.

Features

• 100% local processing (no cloud APIs)
• Type-safe Haskell implementation
• Read text files
• Read from URLs (planned)
• Multilingual support (35+ languages via Piper)
• Language-based voice selection with --lang flag
• Multiple voice support
• Speed control (0.5x - 2.0x)

Setup

1. Piper Binary (Already Done)

The Piper binary is already downloaded in bin/.

2. Voice Models

Installed Voices

• en_US-bryce-medium.onnx (English - US)
• es_ES-davefx-medium.onnx (Spanish - Spain)
• es_MX-claude-high.onnx (Spanish - Mexico)

Preview & Download More Voices

Preview voices: https://rhasspy.github.io/piper-samples/

• Listen to samples of all available voices
• Find voices in 35+ languages (Arabic, Chinese, French, German, Italian, Portuguese, Russian, etc.)

Download voices:

# Example: Download German voice
cd voices
wget https://huggingface.co/rhasspy/piper-voices/resolve/v1.0.0/de/de_DE/thorsten/medium/de_DE-thorsten-medium.onnx
wget https://huggingface.co/rhasspy/piper-voices/resolve/v1.0.0/de/de_DE/thorsten/medium/de_DE-thorsten-medium.onnx.json

URL structure:

https://huggingface.co/rhasspy/piper-voices/resolve/v1.0.0/{language}/{locale}/{voice}/{quality}/{filename}

Examples:

• French: fr/fr_FR/siwis/medium/fr_FR-siwis-medium.onnx
• Italian: it/it_IT/riccardo/x_low/it_IT-riccardo-x_low.onnx
• Portuguese: pt/pt_BR/faber/medium/pt_BR-faber-medium.onnx

Note: Always download both .onnx and .onnx.json files!

3. Build

cd ~/Programs/piper-reader
cabal build

Usage

List available voices

cabal run piper-reader -- list-voices
# Shows: voice name (Language: code)

Read a text file

# Use default voice
cabal run piper-reader -- read-file myfile.txt --play

# Use specific language (auto-selects first matching voice)
cabal run piper-reader -- read-file spanish.txt --lang es --play

# Use specific voice file
cabal run piper-reader -- read-file myfile.txt --voice voices/en_US-bryce-medium.onnx --play

Interactive mode

# English (automatic voice selection)
echo "Hello world" | cabal run piper-reader -- interactive --lang en --play

# Spanish (automatic voice selection)
echo "Hola mundo" | cabal run piper-reader -- interactive --lang es --play

# With custom speed
cabal run piper-reader -- interactive --lang en --speed 1.5 --play
# Type or paste text, then Ctrl+D

Project Structure

piper-reader/
├── bin/           # Piper binary (gitignored)
├── voices/        # Voice models (gitignored)
├── src/           # Haskell source
│   ├── Config.hs
│   ├── CLI.hs
│   ├── TTS/
│   │   ├── Types.hs
│   │   └── Piper.hs
│   └── Content/
│       ├── File.hs
│       ├── Fetch.hs
│       └── Parser.hs
└── app/
    └── Main.hs

Documentation

• USAGE.md - Complete usage guide with examples
• VOICES.md - Voice preview, download instructions, and language reference
• NEXT_STEPS.md - Future enhancements and TODOs

License

MIT