API ReferenceLocal APIs

useTranscribe

React hook — speech-to-text (Whisper) over WebGPU.

Auto-generated from JSDoc in packages/react/src/local/useTranscribe.ts. Do not edit — run npm run docs:api.

useTranscribe

function

function useTranscribe(opts: UseTranscribeOptions): UseTranscribe

UseTranscribe

interface

interface UseTranscribe

`status: TranscribeStatus`

`progress: number`

`text: string`

Text of the most recent transcription.

`error?: Error`

`load: () => Promise<TranscriptionModel>`

`transcribe: (audio: Float32Array, sampleRate?: number) => Promise<string>`

Transcribe a clip (Float32 PCM). Pass the segment audio from useVAD's onSpeechEnd.

UseTranscribeOptions

interface

interface UseTranscribeOptions

`model?: string`

Whisper repo (default onnx-community/whisper-base).

`device?: unknown`

`dtype?: unknown`

`language?: string`

TranscribeStatus

type alias

type TranscribeStatus = idle | loading | ready | transcribing | error

useVAD

React hook — real-time mic voice-activity detection (Silero).

@valora-ai/react

Next Page

On this page

useTranscribe UseTranscribe status: TranscribeStatusprogress: numbertext: stringerror?: Errorload: () => Promise<TranscriptionModel>transcribe: (audio: Float32Array, sampleRate?: number) => Promise<string>UseTranscribeOptions model?: stringdevice?: unknowndtype?: unknownlanguage?: stringTranscribeStatus

Valora is local-first

No API key, no server — everything in this doc runs on-device.