textToSpeech( )
Converts text to speech audio using a loaded TTS model. Supports both streaming and buffered modes.
function textToSpeech(params: { inputType: string; modelId: string; stream: boolean; text: string }, options?: { forceNewConnection?: boolean; profiling?: { enabled?: boolean; includeServerBreakdown?: boolean; mode?: "summary" | "verbose" }; timeout?: number }): { buffer: Promise<number[]>; bufferStream: AsyncGenerator<number>; done: Promise<boolean> }
| Name | Type | Required? | Description |
|---|
params | \{ inputType: string; modelId: string; stream: boolean; text: string \} | ✓ | TTS synthesis parameters |
options | `{ forceNewConnection?: boolean; profiling?: { enabled?: boolean; includeServerBreakdown?: boolean; mode?: "summary" | "verbose" }; timeout?: number }` | ✗ |
| Field | Type | Required? | Description |
|---|
inputType | string | ✓ | The type of input content, typically "text" |
modelId | string | ✓ | The ID of the loaded TTS model to use for synthesis |
stream | boolean | ✓ | When true, audio bytes are emitted incrementally via bufferStream. When false, the full audio is returned via buffer |
text | string | ✓ | The text content to synthesize into speech |
{ buffer: Promise<number[]>; bufferStream: AsyncGenerator<number>; done: Promise<boolean> }
| Field | Type | Description |
|---|
buffer | Promise | Resolves with the complete audio buffer as a byte array when stream is false. Resolves to an empty array in streaming mode |
bufferStream | AsyncGenerator | Yields audio bytes incrementally in streaming mode. Empty generator when stream is false |
done | Promise | Resolves to true when synthesis is complete |
import { textToSpeech } from "@qvac/sdk";
// Streaming mode — process audio chunks as they arrive
const { bufferStream, done } = textToSpeech({
modelId: "my-tts-model",
text: "Hello, welcome to QVAC!",
inputType: "text",
stream: true,
});
const audioChunks = [];
for await (const byte of bufferStream) {
audioChunks.push(byte);
}
await done;
// Buffered mode — wait for the full audio
const { buffer } = textToSpeech({
modelId: "my-tts-model",
text: "Hello, welcome to QVAC!",
inputType: "text",
stream: false,
});
const audioBytes = await buffer;
console.log("Audio size:", audioBytes.length, "bytes");
