PlayAI Text-to-Speech API Playground

Language

Voice

Model

Temporary API Latency Update

Due to high demand from our recent launch, our real-time conversational API may experience brief delays. We're working to optimize performance and appreciate your patience.

Meet Play 3.0 mini

For developers getting started with integrating PlayAI voice models in their apps.

Real-Time Communication with WebSockets

For developers seeking lowest latency conversational AI experiences, learn how to integrate with websockets.

Contact Sales

Looking for a custom solution or have questions about our API? Contact our sales team.

Generated Audio Stats

Inference Time:—

Time to First Audio:(Inference + Network)

—

Audio Generation Time:—

WebSocket Status:Disconnected

Advanced Voice Controls

Sample Rate

Choose the audio sample rate

Custom Finetune Voice ID

Provide a custom finetune voice ID (optional)

Speed: 1

Adjust the speed of speech (0.1-5.0)

Temperature: 0.7

Control randomness in generation (0.1-2.0)

Top P: 0.35

Nucleus sampling probability (0.0-1.0)

Repetition Penalty: 3

Penalize repetition in generation (0.0-10.0)

Length Penalty: 1.1

Adjust output length (0.1-2.0)

Voice Guidance: 0

Control voice adherence (0.0-2.0). Default is 0.

Seed

Set a seed for reproducible results (optional)

Style Guidance: 0.5

Control style adherence (0.1-10.0)