Voice API

Build voice into anything

A streaming text-to-speech API engineered for low latency, high concurrency and clean integration.

No subscriptions · 1 credit = $0.10 · Credits never expire

<300ms
Time to first audio
99.95%
Uptime SLA
5
Official SDKs
200+
Voices via API
tts-stream.ts
import { ITSPACE24 } from "@vocalix/sdk";

const vocalix = new ITSPACE24(process.env.VOCALIX_KEY);

// Real-time streaming TTS — playback starts in <300ms
const stream = await vocalix.tts.stream({
  voice: "nova",
  input: "Hello from ITSPACE24.",
  format: "mp3",
});

for await (const chunk of stream) {
  speaker.write(chunk);
}
Response
Sub-second time-to-first-audio, chunked over HTTP/2 or WebSocket.
POST/v1/ttsOne-shot generation
GET/v1/voicesList 200+ voices
POST/v1/voices/cloneCreate a clone
WS/v1/streamBidirectional streaming
Developer dashboard

Everything a builder needs in one place

Live API reference, request inspector, audio results and usage analytics — built for developers, not just docs readers.

Production-ready, day one

Sub-300ms latency

Designed for real-time conversation UX and voice agents.

Streaming output

Start playback before the sentence is finished generating.

WebSocket support

Full bidirectional streaming for agents and live interfaces.

SDKs included

TypeScript, Python, Go, Swift and Kotlin — typed and tested.

99.95% SLA

Multi-region failover, observable from day one in your dashboard.

Volume pricing

Scale to billions of characters per month with predictable cost.

Loved by creators & teams

From indie podcasters to Fortune 500 product teams.

We swapped our previous TTS provider for ITSPACE24 and our voice-agent p99 latency dropped by 40%.
Sven Olsen
Staff Engineer, Orbital
The SDKs are the cleanest I've used in this category.
Devi Krishnan
Founding engineer
Streaming over WebSocket just works. No babysitting.
Marc Lavoie
CTO, Foundry

Frequently asked questions

How is the API priced?+

Same credit system as the apps. 1 credit ≈ 1,000 characters. Volume discounts kick in automatically above 500K credits per month.

What's the latency?+

Time-to-first-audio is typically 180–280ms in production, including TTS + network.

Do you support streaming?+

Yes — both HTTP/2 chunked and WebSocket bidirectional. SDKs handle reconnection.

Is there a free tier?+

Buy a 50-credit pack to build and test the API before scaling up.

Ship voice this week

Buy 50 credits and grab an API key from your dashboard.

Credits never expire · No subscription required