Leopard Speech-to-Text
React API

API Reference for the Leopard React SDK (npmjs).

useLeopard()

function useLeopard(): {
  result: LeopardTranscript | null;
  isLoaded: boolean;
  error: Error | null;
  init: (
    accessKey: string,
    model: LeopardModel,
    options?: LeopardOptions,
  ) => Promise<void>;
  processFile: (file: File) => Promise<void>;
  startRecording: (maxRecordingSec?: number) => Promise<void>;
  stopRecording: () => Promise<void>;
  isRecording: boolean;
  recordingElapsedSec: number;
  release: () => Promise<void>;
}

React Hook for Leopard speech-to-text engine.

Returns

• result LeopardTranscript | null : A state containing the transcript result value.
• isLoaded boolean : A state indicating whether the engine has initialized successfully.
• error Error | null : An Error object containing details about the error.
• init (...) => Promise<void> : A method to initialize Leopard given the arguments. See init.
• processFile (...) => Promise<void> : A method to process audio files. See processFile.
• startRecording (...) => Promise<void> : A method to start recording audio. See startRecording.
• stopRecording () => Promise<void> : A method to stop recording audio and begin processing. See stopRecording.
• isRecording boolean : A state indicating whether webVoiceProcessor is passing audio to the engine.
• recordingElapsedSec number : The amount of time (in seconds) elapsed while recording.
• release () => Promise<void> : Releases resources acquired by useLeopard. See release.

init()

async function init(
  accessKey: string,
  model: LeopardModel,
  options: LeopardOptions = {},
): Promise<void>

Initializes useLeopard hook. Sets isLoaded to true.

Parameters

• accessKey string : AccessKey obtained from Picovoice Console.
• model LeopardModel : Object containing Leopard model.
• options LeopardOptions : Optional configuration arguments.

processFile()

async function processFile(
  file: File
): Promise<void>

Processes an audio file, which should be provided as a File object with a MIME type of audio and encoded in linear PCM format.

Parameters

• file File : File object with the MIME type = audio.

startRecording()

async function startRecording(
  maxRecordingSec: number = 120
): Promise<void>

Start recording audio. If started, isRecording is set to true.

Parameters

• maxRecordingSec number : Optional configuration argument to set the maximum time in seconds Leopard is able to record. If unspecified, a default value of 120 will be used. This is to prevent unbounded memory usage.

stopRecording()

async function stopRecording(): Promise<void>

Stop recording audio and begin processing the buffered audio. Sets isRecording to false.

release()

function release(): Promise<void>

Releases resources acquired by the useLeopard. Sets isLoaded and isRecording to false, recordingElapsedSec to 0, and error to null.