Orpheus-3b-Kaya-Q8_0 Open-source Text-to-Speech Model - Free Generation of 24kHz English Audio

Orpheus 3b Kaya Q8 0.gguf

Developed by lex-au

An 8-bit quantized text-to-speech model fine-tuned from Canopy Labs' pre-trained model, supporting 24kHz English audio generation

Downloads 38

Release Time : 4/19/2025

Model Overview

This is a 3-billion-parameter model optimized for speech synthesis tasks, using GGUF Q8_0 quantization format for efficient inference

Efficient quantization

Uses 8-bit GGUF (Q8_0) quantization format to significantly reduce model size and computational requirements while maintaining quality

High sample rate audio

Supports generating 24kHz mono audio for higher-quality speech output

Fast inference

Optimized for use with efficient inference servers like Orpheus-FastAPI

English text-to-speech

24kHz audio generation

Quantized model inference

Speech synthesis applications

Voice assistants

Provides natural voice output for smart assistants

24kHz high-quality speech

Audiobook generation

Automatically converts text content to speech

Property	Details
Model Type	Text-to-Speech (TTS)
Architecture	Token-to-audio language model
Parameters	~3 billion
Quantisation	8-bit GGUF (Q8_0)
Sampling Rate	24kHz mono
Training Epochs	1
Training Data	lex-au/Orpheus-3b-Kaya
Languages	English

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base