tasks-audio package {#tasks_audio_package}

Classes

Class	Description
AudioClassifier	Performs audio classification.
AudioTranscriber	MediaPipe Audio Task that performs automatic speech recognition (ASR) in unary and streaming modes via WASM SIMD or WebGPU.
FilesetResolver	Resolves the files required for the MediaPipe Task APIs. The returned filesets require that the Wasm files are published without renaming. If this is not possible, you can invoke the MediaPipe Tasks APIs using a manually created WasmFileset.
TextToSpeech	MediaPipe Audio Task that performs neural text-to-speech synthesis in unary and streaming modes via WASM SIMD or WebGPU.
TextToSpeechStream	Active streaming synthesis session returned by TextToSpeech.synthesizeStream(). Supports both pull-based readNext() and for await (const chunk of stream) iteration, as well as incremental text input via pushText() and finishInput().

Interfaces

Interface	Description
AudioClassifierOptions	Options to configure the MediaPipe Audio Classifier Task
AudioClassifierResult	Classification results of a model.
AudioTranscriberOptions	Options to configure the MediaPipe AudioTranscriber Task on the Web.
AudioTranscriberResult	Result of an AudioTranscriber invocation.
Category	A classification category.
Classifications	Classification results for a given classifier head.
SynthesisOptions	Per-utterance synthesis overrides passed to TextToSpeech.synthesize() and TextToSpeech.synthesizeStream().
TextToSpeechAudioChunk	A single streaming PCM audio chunk emitted by TextToSpeechStream.
TextToSpeechOptions	Options to configure the MediaPipe TextToSpeech Task on the Web.
TextToSpeechResult	Result of a unary TextToSpeech.synthesize() invocation.
TranscriptionSegment	A single transcribed segment emitted by AudioTranscriber.
VoiceInfo	Voice metadata descriptor returned by TextToSpeech.listVoices().
WordTimestamp	Word-level timing alignment mapping an input word to audio seconds.

Type Aliases

Type Alias	Description
AudioTranscriberModelArchitecture	Model architecture family for AudioTranscriber.

AudioTranscriberModelArchitecture {#audiotranscribermodelarchitecture}

Model architecture family for AudioTranscriber.

Signature:

export declare type AudioTranscriberModelArchitecture = "AUTO" | "OMNI" | "TINY_GEMMA";