Interface AudioTranscriptionCapability<TInput, TOutput>

Provider-agnostic audio transcription capability.

Converts speech audio to text/transcript artifacts.

interface AudioTranscriptionCapability<
    TInput = unknown,
    TOutput = NormalizedChatMessage[],
> {
    transcribeAudio(
        request: AIRequest<TInput>,
        ctx: MultiModalExecutionContext,
        signal?: AbortSignal,
    ): Promise<AIResponse<TOutput>>;
}

Type Parameters

TInput = unknown
TOutput = NormalizedChatMessage[]

Hierarchy (View Summary)

ProviderCapability
- AudioTranscriptionCapability

Implemented by

Index

Methods

transcribeAudio

Methods

transcribeAudio

transcribeAudio(
    request: AIRequest<TInput>,
    ctx: MultiModalExecutionContext,
    signal?: AbortSignal,
): Promise<AIResponse<TOutput>>
Transcribe input audio into text.
Parameters
- request: AIRequest<TInput>
  Unified AI request containing transcription input.
- ctx: MultiModalExecutionContext
  Execution context.
- Optionalsignal: AbortSignal
  Optional abort signal.
Returns Promise<AIResponse<TOutput>>
AIResponse containing normalized transcript artifacts.
- Defined in src/core/provider/capabilities/AudioCapability.ts:31