5-minute setup

VibeVoice API Quickstart

Generate your first studio-grade voice clip without leaving the console. Follow the exact curl + Python commands the team uses for smoke tests.

Request API Key Jump to Quickstart

Latency

<180 ms

Streaming endpoint average.

Requests

1M / day

Enterprise throttling ceiling.

Regions

7

US, EU & APAC data residency.

SDKs

Python · JS

CLI-friendly helpers.

Pricing & limits

Choose your plan

Talk to sales
Plan Monthly price Included minutes Rate limit Highlights
Free Builder $0 60 minutes 20 req/min AWS-compatible key management, sandbox watermark.
Pro Studio $79 800 minutes 120 req/min Streaming endpoint, priority queue, custom voices.
Enterprise Custom Unlimited 1k req/min* Dedicated region, SLA 99.9%, governance toolkit.

*Soft limit. Can be extended via signed capacity agreement.

Hands-on

Quickstart steps

  1. Step 1

    Create an API key

    Visit the VibeVoice console → API Keys → New key → enable voice.synthesis scope and copy the token. Store it in your secrets manager.

  2. Step 2

    Send a cURL request

    Use the REST endpoint to generate an MP3 clip. Replace $API_KEY with your token.

    curl https://api.vibevoice.online/v1/speech \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "voice": "berlin_pro",
    "language": "de-DE",
    "format": "mp3",
    "text": "Willkommen zur VibeVoice API Quickstart Demo."
  }' --output sample.mp3

  3. Step 3

    Install the Python SDK

    The SDK wraps retries, streaming, and file helpers.

    pip install vibevoice-sdk

from vibevoice import SpeechClient

client = SpeechClient(api_key="${API_KEY}")

result = client.generate(
    text="Welcome to VibeVoice",
    voice="london_narrator",
    language="en-US",
    output_format="wav",
)

with open("welcome.wav", "wb") as f:
    f.write(result.audio)

Sample response

Successful requests return metadata you can log or audit.

{
  "request_id": "req_8fd1d7",
  "status": "succeeded",
  "voice": "berlin_pro",
  "duration_ms": 4180,
  "sample_rate": 48000,
  "format": "mp3",
  "audio": "base64..."
}

Decode the Base64 payload or use the SDK helper result.save("file.mp3").

SDK + tooling

  • Python: vibevoice-sdk with async streaming, caching, SSML helpers.
  • Node: REST wrapper via @vibevoice/client (beta).
  • Postman: Import the official collection from the console.
  • CLI: npx vibevoice-cli synth --text \"demo\" for smoke tests.
GitHub SDK Postman collection

Deploy

Serverless ready

Use Azure Functions or Vercel Edge to proxy requests, inject secrets, and log usage centrally.

Observe

Tracing hooks

Forward request_id to Datadog or OpenTelemetry for latency distributions and retry alerts.

Secure

Fine-grained scopes

Issue per-environment keys, limit to read/write operations, and rotate via API for zero downtime.

FAQ

Do I need a credit card for the Free Builder plan?

No. Billing details are only required when you upgrade to Pro Studio.

Can I store audio outside VibeVoice?

Yes. Responses contain raw audio buffers so you can persist them in S3, Azure Blob, or attach to CMS assets.

How do I handle long-form projects?

Use batch mode with chunk_strategy=\"smart\" to split scripts, or enable streaming for sub-200 ms playback.

Is on-prem deployment supported?

Enterprise tier unlocks private container registries so you can host inference inside your own VNet.

Need deeper coverage?

Jump back to the full API reference or talk to our solution engineers for compliance reviews.

Full endpoint list Book a 30-min session

Start building

Run your first VibeVoice API call today

Sign in, grab a key, and ship production-ready voices with audited latency and enterprise security.

Generate API key Follow quickstart