Unlimited multilingual TTS for web and Node.js at a predictable low cost

Browser quickstart

Activate a paid plan, create a project, register your browser origins, then initialize the package with its publishable key.

Install the public package from npm.

npm install @cheap-tts/browser-sdk

import { CheapTts } from "@cheap-tts/browser-sdk";

const tts = new CheapTts({
  publishableKey: "pk_live_your_project",
});

const result = await tts.synthesize({
  text: "Xin chào từ Cheap TTS.",
  voice: "vi-entertainment",
});

audio.src = URL.createObjectURL(result.audio);
await audio.play();

Open the live browser example →

Generate voice in your backend

Cheap TTS gives backend applications unlimited generation, the complete curated voice catalog, and a language-aware pipeline at a predictable low cost. Reusable local assets keep repeat generation efficient and fast.

Install the public package from npm.

npm install @cheap-tts/node

import { CheapTts } from "@cheap-tts/node";

const tts = new CheapTts({
  secretKey: process.env.CHEAP_TTS_SECRET_KEY,
});

const result = await tts.synthesize({
  text: "Xin chào từ Cheap TTS.",
  voice: "vi-entertainment",
});

await result.save("./audio.wav");
await tts.destroy();

Warm up a voice

Call loadVoice() when a worker starts so the first generation request is ready sooner. The SDK automatically manages reusable voice assets in its local cache.

await tts.loadVoice({
  voice: "en-narration",
  onProgress: ({ progress }) => console.log(progress),
});

Node API reference

Discover voices

Use stable voice IDs to select a language, voice style, and use case.

const voices = await tts.voices();

// Filter the curated voices for one language
const vietnamese = voices.filter((voice) => voice.language === "vi");

Show browser progress

Preload the selected voice before enabling generation. The SDK handles voice loading and reuses it on later visits when browser storage is available.

await tts.loadVoice({
  voice: "en-narration",
  onProgress: ({ progress }) => {
    console.log(progress); // 0..1, or null
  },
});

await tts.synthesize({
  text,
  voice: "en-narration",
  rate: 1.1,
  onProgress: ({ stage, progress }) => {
    console.log(stage, progress); // 0..1, or null
  },
});

Browser API reference

new CheapTts(options)

tts.voices()

Returns the enabled voice catalog as { id, name, language, useCase }[].

tts.synthesize(options)

Returns { audio, arrayBuffer, seconds, voice }. audio is a browser-ready audio/wav Blob.

tts.destroy()

Terminates the worker and rejects active jobs. Call this when permanently unmounting the integration.

Browser access security

A publishable key is expected to be visible in frontend code. It is not a secret and does not grant access by itself.

• The key is accepted only from origins trusted for your project.
• An active paid subscription is checked when sessions are created and text is prepared.
• The SDK exchanges it for a signed, short-lived browser session.
• Text preparation requires that session and is rate limited.
• Never place a Node or server secret key in browser code.

Common errors