Lex-au/Orpheus-FastAPI
A high-performance Text-to-Speech server using deep learning models with GPU optimization and multilingual voice support.
Velocity · 7d
+1.6
★ / day
Trend
→steady
star history
This repository provides a FastAPI-based TTS inference server that runs quantized Orpheus TTS models optimized for RTX GPUs. It exposes an OpenAI-compatible API endpoint and includes a web UI with 24 voice actors across 8 languages, emotion tagging support, and language-specific fine-tuned model variants available on HuggingFace.