What it does
Core capabilities at a glance
- Multi-language STT (English, Chinese, Japanese, and more)
- Multi-accelerator support (CUDA, CoreML, OpenVINO)
- Voice activity detection (VAD) included
- TTS with multiple voices
- Speaker recognition and diarization
- WebAssembly build for browser
Deep dive
The full breakdown - performance, comparisons, and setup
Sherpa Onnx
Sherpa Onnx is the swiss-army knife of speech processing. It bundles speech-to-text, text-to-speech, voice activity detection, speaker recognition, and language identification into a single cross-platform library.
What it is
Sherpa Onnx is an open-source speech processing toolkit built on ONNX Runtime. It supports multiple speech tasks through a unified API and runs on CPU, CUDA, CoreML, and OpenVINO.
How it stacks up
| Sherpa Onnx | Whisper.cpp | Piper TTS | Kokoro TTS | |
|---|---|---|---|---|
| STT + TTS | ✓ | STT only | TTS only | TTS only |
| VAD | ✓ | Via lib | ✗ | ✗ |
| Speaker ID | ✓ | ✗ | ✗ | ✗ |
| WebAssembly | ✓ | ✗ | ✗ | ✗ |
| Best for | All-in-one | STT quality | Home Assistant | Voice quality |
Get started
import sherpa_onnx
recognizer = sherpa_onnx.OfflineRecognizer.from_models(
encoder="sherpa-onnx-whisper-tiny.en",
)
result = recognizer.transcribe("audio.wav")
print(result.text)When to use something else
- Best STT accuracy: Whisper.cpp
- Best TTS quality: Kokoro TTS
- Home Assistant integration: Piper TTS
Frequently asked
Quick answers to common questions
What is Sherpa Onnx?
Sherpa Onnx is a tts-stt tool for local AI workloads. Cross-platform speech processing toolkit for STT, TTS, VAD, speaker recognition, and language ID.
Is Sherpa Onnx free and open source?
Yes, Sherpa Onnx has 12,797 GitHub stars and is licensed under Apache-2.0. You can self-host it for free on linux, windows, macos, docker, web.
What platforms does Sherpa Onnx support?
Sherpa Onnx runs on linux, windows, macos, docker, web.
What hardware do I need for Sherpa Onnx?
The hardware requirements depend on which models you run. Check our hardware directory for compatible GPUs and systems. Sherpa Onnx has 12,797 GitHub stars and an active community.
Does Sherpa Onnx support GPU acceleration?
Sherpa Onnx supports GPU acceleration via CUDA, Metal, or Vulkan depending on your platform. For the best performance, pair it with an NVIDIA RTX 4090 or 5090.
What are the best alternatives to Sherpa Onnx?
Popular alternatives include other tts-stt tools in our directory. Browse our full collection at /tool for comparisons, community reviews, and benchmark data to find the right fit for your workflow.
How much does Sherpa Onnx cost?
Sherpa Onnx is free-open-source. It is completely free and open source to self-host.
Pairs well with
Complementary tools, models, and hardware
Comments coming soon
Configure NEXT_PUBLIC_GISCUS_REPO_ID and NEXT_PUBLIC_GISCUS_CATEGORY_ID at giscus.app to enable.