Options to configure the MediaPipe AudioTranscriber Task on the Web.
Signature:
export declare interface AudioTranscriberOptions extends TaskRunnerOptions
Extends: TaskRunnerOptions
Properties
| Property | Type | Description |
|---|---|---|
| enableInterimResults | boolean | Whether to emit interim (non-final) transcription segments during streaming. Defaults to false. |
| modelArchitecture | AudioTranscriberModelArchitecture | number | Model architecture backend family ('AUTO', 'OMNI', or 'TINY_GEMMA', or numeric code 0=AUTO, 1=OMNI, 2=TINY_GEMMA). Defaults to 'AUTO'. |
| modelAssetSize | number | Optional byte length of baseOptions.modelAssetBuffer when provided as a ReadableStreamDefaultReader. |
| prompt | string | Optional system or user prompt override for generative ASR models. |
| sampleRateHz | number | Input audio sample rate in Hz. Defaults to 16000 Hz when unset or 0. |
AudioTranscriberOptions.enableInterimResults
Whether to emit interim (non-final) transcription segments during streaming. Defaults to false.
Signature:
enableInterimResults?: boolean;
AudioTranscriberOptions.modelArchitecture
Model architecture backend family ('AUTO', 'OMNI', or 'TINY_GEMMA', or numeric code 0=AUTO, 1=OMNI, 2=TINY_GEMMA). Defaults to 'AUTO'.
Signature:
modelArchitecture?: AudioTranscriberModelArchitecture | number;
AudioTranscriberOptions.modelAssetSize
Unused: the model is always written to the Wasm virtual filesystem, which the native AudioTranscriber requires.
Optional byte length of baseOptions.modelAssetBuffer when provided as a ReadableStreamDefaultReader.
Signature:
modelAssetSize?: number;
AudioTranscriberOptions.prompt
Optional system or user prompt override for generative ASR models.
Signature:
prompt?: string;
AudioTranscriberOptions.sampleRateHz
Input audio sample rate in Hz. Defaults to 16000 Hz when unset or 0.
Signature:
sampleRateHz?: number;