Now available — ultra-low latency streaming

ship voice in
five lines of code

Production-grade text-to-speech with low-latency streaming and 9 languages. Usage-based pricing starting free.

Start for free View documentation

Try it

Hear the difference

Type anything and listen in real time. No sign-up required.

114 / 200

Pricing

Simple, usage-based pricing

Start free, scale as you grow. No hidden fees, no commitments.

Free

Get started with voice AI at no cost.

$0forever

10,000 chars/month

2 custom voices
All endpoints included
Hard capped — no surprise bills

Start for free

Developer

For apps moving to production.

$10/mo

500,000 chars/month included

10 custom voices
$8 per 1M chars overage
Priority support

Get started

Growth

For high-volume production workloads.

$25/mo

2,000,000 chars/month included

Unlimited custom voices
$6 per 1M chars overage
Priority support

Get started

Developer-first

Ship voice features
in minutes

A clean, predictable REST API and WebSocket streaming. OpenAI-compatible endpoint, streaming out of the box, and comprehensive docs that actually help.

Get API key Read the docs

speak.py
 1import httpx
 2
 3client = httpx.Client(
 4    base_url="https://api.phemius.dev",
 5    headers={"Authorization": "Bearer ph_live_..."}
 6)
 7
 8response = client.post("/v1/speech", json={
 9    "text": "Ship voice in five lines of code.",
10    "voice": "aria",
11    "model": "phemius-fast",
12    "speed": 1.0
13})
14
15with open("output.pcm", "wb") as f:
16    f.write(response.content)

ship voice infive lines of code