Core
Audio Generation
Provider-neutral text-to-speech model contracts and request builders.
Import from @anvia/core or @anvia/core/audio-generation.
AudioGenerationModel
interface AudioGenerationModel<RawResponse = unknown> {
readonly provider?: string;
readonly defaultModel?: string;
audioGeneration(request: AudioGenerationRequest): Promise<AudioGenerationResponse<RawResponse>>;
}Purpose: provider-neutral audio generation and text-to-speech contract.
Return behavior: providers return generated audio bytes, optional mediaType, and the provider response as rawResponse.
Request Builder
type AudioGenerationRequest = {
text: string;
voice: string;
speed: number;
additionalParams?: JsonValue;
};
type AudioGenerationResponse<RawResponse = unknown> = {
audio: Uint8Array;
mediaType?: string;
rawResponse: RawResponse;
};
class AudioGenerationRequestBuilder<Model extends AudioGenerationModel = AudioGenerationModel> {
text(text: string): this;
voice(voice: string): this;
speed(speed: number): this;
additionalParams(additionalParams: JsonValue): this;
build(): AudioGenerationRequest;
send(): Promise<AudioGenerationResponse>;
}
const response = await audioGenerationRequest(model)
.text("Welcome to Anvia.")
.voice("alloy")
.speed(1)
.additionalParams({ response_format: "mp3" })
.send();Purpose: Chainable builder for audio generation requests.
Defaults: text: "", voice: "", and speed: 1.
Notable errors: provider adapters reject on SDK or provider errors.