Pre-converted weights for running Qwen3-TTS 1.7B entirely via llama.cpp + ONNX Runtime. No PyTorch required.
Browse all Text To Speech models →