Text to Speech

Overview

The Text to Speech API turns input text into natural, human-like audio in a voice of your choice. Pick any voice from the library or one of your cloned voices, select a model, and request an audio stream or file.

Quickstart

Install the official SDK, then convert text to speech with a voice ID and a model. The example below returns MP3 audio from the multilingual model.

import { IntuneLabsClient } from "@intunelabs/intunelabs-js";

const client = new IntuneLabsClient({ apiKey: process.env.INTUNELABS_API_KEY });

const audio = await client.textToSpeech.convert("JBFqnCBsd6RMkjVDRZzb", {
  text: "The first move is what sets everything in motion.",
  modelId: "intune_multilingual_v2",
  outputFormat: "mp3_44100_128",
});

Pass your API key through an environment variable. Never hardcode secrets in client-side code or commit them to source control.

Supported models

• intune_v3 — the most expressive output, ideal for performance and multi-speaker dialogue.
• intune_multilingual_v2 — stable, consistent quality for long-form narration.
• intune_flash_v2_5 — ultra-low latency for real-time use.

Best practices

• Add punctuation and natural phrasing to guide pacing and intonation.
• Stream audio for interactive experiences to minimize time-to-first-byte.
• Cache repeated generations to save credits and reduce latency.
• Test voices against your content before committing to a production voice.