react-native-sherpa-onnx and react-native-speech-recognition-kit
Both tools provide speech recognition capabilities for React Native, but tool A offers a more comprehensive suite of offline on-device speech processing features (STT/TTS/Diarization/VAD) using Sherpa-ONNX, while tool B provides a simpler Turbomodule to access platform-native speech recognition; thus, they are competitors, with tool A being a more feature-rich and self-contained solution.
About react-native-sherpa-onnx
XDcobra/react-native-sherpa-onnx
React Native TurboModule for Sherpa-ONNX offline on-device Speech Processing (STT/TTS/Diarization/VAD) completely offline on the device. Support for Android & iOS
Built on TurboModules for native performance, it supports streaming speech recognition with partial results and endpoint detection, plus batch and streaming TTS with multiple model architectures (Zipformer, Paraformer, Whisper, VITS, Matcha, Kokoro). The SDK bundles sherpa-onnx 1.12.31 with automatic iOS XCFramework downloads and flexible Android execution providers (CPU, NNAPI, XNNPACK, QNN), while an integrated download manager handles runtime model fetching with background support and Play Asset Delivery integration on Android.
About react-native-speech-recognition-kit
Gautham495/react-native-speech-recognition-kit
React Native Turbo Module to access Speech Recognition in Android & iOS
Leverages native speech engines (Apple's SFSpeechRecognizer and Google's SpeechRecognizer) for on-device, offline transcription with real-time partial results and voice level detection across 50+ languages. Built as a TurboModule for synchronous native bridge performance, supporting multi-language recognition and working with Expo Dev Client / Prebuild without requiring cloud APIs.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work