sherpa-onnx-tts

You can generate speech locally without cloud services by using the sherpa-onnx offline CLI to convert text to speech. This preserves privacy and reduces latency by running entirely on your machine.

Offline TTS requires setting SHERPA_ONNX_RUNTIME_DIR and SHERPA_ONNX_MODEL_DIR. These variables must point to the downloaded sherpa-onnx runtime and the Piper en_US lessac model directory, respectively.

Yes, local text to speech via sherpa-onnx works across macOS, Linux, and Windows. It supports workflows needing offline audio generation for applications, accessibility tools, and embedded voice features on these platforms.

No, offline TTS does not require an internet connection to synthesize audio. Once the runtime and Piper en_US lessac model are downloaded, the CLI generates speech entirely locally without relying on cloud services.

The best way to add voice feedback to offline kiosks is using the sherpa-onnx local TTS wrapper. It provides simple CLI-based speech synthesis for embedded voice output in environments lacking internet connectivity.

Use local TTS instead of cloud-based speech synthesis to preserve privacy and reduce audio generation latency. Running the sherpa-onnx CLI offline avoids sending text data to external cloud services.