foxl-ai sherpa-onnx-tts
Local offline TTS using sherpa-onnx.
Setup
- Download runtime for your OS from sherpa-onnx releases
- Download a voice model (e.g., vits-piper-en_US-lessac-high)
- Set environment variables:
export SHERPA_ONNX_RUNTIME_DIR=~/.pilot/tools/sherpa-onnx-tts/runtime
export SHERPA_ONNX_MODEL_DIR=~/.pilot/tools/sherpa-onnx-tts/models/vits-piper-en_US-lessac-high
Usage
sherpa-onnx-tts -o ./output.wav "Hello from local TTS."
Options
-o/--output- Output WAV file--model-file- Specific model file if multiple exist--tokens-file- Override tokens file--data-dir- Override data directory
Notes
- Completely offline, no API keys needed
- Pick different models from sherpa-onnx tts-models releases
- Supports multiple languages depending on model