Speak through ElevenLabs, OpenAI, Google, Amazon, and more, with a single integration. Pay only for the characters you synthesize.
No credit card required · 50,000 free characters every month
Route to any of these engines from one endpoint
Drop in the SDK and get every provider behind one schema. Change thevoiceormodelfield to switch engines, no other code changes required.
import { TTSGateway } from "ttsgateway";
const tts = new TTSGateway({ apiKey: process.env.TTS_GATEWAY_KEY });
const audio = await tts.speak({
text: "One API for every text-to-speech engine.",
voice: "aria",
model: "auto", // routes to the fastest healthy engine
format: "mp3",
});Everything you need to ship reliable, cost-efficient speech, without babysitting provider status pages.
Swap providers with a single field. No rewrites, no vendor lock-in, no separate SDKs to learn.
A provider goes down or rate-limits you, requests reroute to the next healthy engine automatically.
Repeated phrases get served from cache instantly. Route by cost, latency, or voice quality per request.
Track characters synthesized, spend, and latency per provider, per voice, per API key, in real time.
Use your existing provider accounts, or let usage run through ours. Your audio, your data, never stored.
Low-latency chunked audio streaming for real-time voice agents, over the same unified endpoint.
No subscriptions, no seats, no monthly minimums. Top up a balance and spend it down as you synthesize speech.
Fast, natural voices for everyday use
ElevenLabs, PlayHT, Cartesia and other top-tier engines
Identical text served from cache
Get an API key in seconds and start streaming speech from any engine.