Expand description
tauri-plugin-whisper-core — Offline, real-time Speech-to-Text for Tauri v2.
This plugin wraps whisper.cpp
(via [whisper-rs]) to provide streaming transcription with hardware
acceleration (CUDA / Metal / CoreML), Silero VAD, and Pyannote Speaker Diarization.
§Quick Start
ⓘ
tauri::Builder::default()
.plugin(tauri_plugin_whisper_core::init())
.run(tauri::generate_context!())
.expect("error while running tauri application");Re-exports§
pub use state::AudioChannel;pub use state::AudioContext;pub use state::DiarizedResultPayload;pub use state::DiarizedSegmentPayload;pub use state::DownloadProgressPayload;pub use state::FinalResultPayload;pub use state::PartialResultPayload;pub use state::SegmentPayload;pub use state::WhisperPluginState;pub use state::RecordingTranscribeProgressPayload;
Modules§
- sck
- ScreenCaptureKit audio capture backend.
- state
- vad
- Voice Activity Detection (VAD) using Silero VAD (ONNX) or fallback energy detector.
Structs§
- Recording
Transcript - The complete transcript returned by
transcribe_recording. - Recording
Transcript Segment - A single segment in the final recording transcript.
- Transcript
Segment - A single transcribed segment with timestamps.
Enums§
- Whisper
Error - All errors that the whisper plugin can surface to the frontend.
Traits§
- AsrBackend
- Pluggable backend for speech-to-text inference.
Functions§
- init
- Initialises the whisper-core plugin.