Orpheus FASTAPI | Advanced Text-to-Speech

Generate Speech

Text to speak

0 / 8192 characters

Language

English

Voice

Shaurya

Advanced options

Model

Speed

Slower 1.0 Faster

These settings will be saved to a .env file. Restart the server to apply changes.

API URL

API Timeout (seconds)

Max Tokens

Temperature

Top P

Repetition Penalty Fixed at 1.1

Value hardcoded to 1.1 for optimal generation quality

Server Host

Server Port

Supports emotion tags: <laugh>, <sigh>, etc.

Use <laugh> to add laughter to the speech
Use <sigh> for a sighing sound
Other supported tags: <chuckle>, <cough>, <sniffle>, <groan>, <yawn>, <gasp>
For longer audio, the system can generate up to 2 minutes of speech in a single request
For API access, use the /v1/audio/speech endpoint (OpenAI compatible)