zai-org / GLM-TTS

GLM-TTS: Controllable & Emotion-Expressive Zero-shot TTS

💻 GitHub Repository | 🤗 Online Demo | 🛠️ Audio.Z.AI

📖 Model Introduction

GLM-TTS is a high-quality text-to-speech (TTS) synthesis system based on large language models, supporting zero-shot voice cloning and streaming inference. The system adopts a two-stage architecture combining an LLM for speech token generation and a Flow Matching model for waveform synthesis.

By introducing a Multi-Reward Reinforcement Learning framework, GLM-TTS significantly improves the expressiveness of generated speech, achieving more natural emotional control compared to traditional TTS systems.

Key Features

• Zero-shot Voice Cloning: Clone any speaker's voice with just 3-10 seconds of prompt audio.
• RL-enhanced Emotion Control: Utilizes a multi-reward reinforcement learning framework (GRPO) to optimize prosody and emotion.
• High-quality Synthesis: Generates speech comparable to commercial systems with reduced Character Error Rate (CER).
• Phoneme-level Control: Supports "Hybrid Phoneme + Text" input for precise pronunciation control (e.g., polyphones).
• Streaming Inference: Supports real-time audio generation suitable for interactive applications.
• Bilingual Support: Optimized for Chinese and English mixed text.

System Architecture

GLM-TTS follows a two-stage design:

1. Stage 1 (LLM): A Llama-based model converts input text into speech token sequences.
2. Stage 2 (Flow Matching): A Flow model converts token sequences into high-quality mel-spectrograms, which are then turned into waveforms by a vocoder.

Reinforcement Learning Alignment

To tackle flat emotional expression, GLM-TTS uses a Group Relative Policy Optimization (GRPO) algorithm with multiple reward functions (Similarity, CER, Emotion, Laughter) to align the LLM's generation strategy.

Evaluation Results

Evaluated on seed-tts-eval . GLM-TTS_RL achieves the lowest Character Error Rate (CER) while maintaining high speaker similarity.

Quick Start

Installation

git clone [https://github.com/zai-org/GLM-TTS.git](https://github.com/zai-org/GLM-TTS.git)
cd GLM-TTS
pip install -r requirements.txt

Command Line Inference

python glmtts_inference.py \
    --data=example_zh \
    --exp_name=_test \
    --use_cache \
    # --phoneme # Add this flag to enable phoneme capabilities.

Shell Script Inference

bash glmtts_inference.sh

Acknowledgments & Citation

We thank the following open-source projects for their support:

• CosyVoice - Providing frontend processing framework and high-quality vocoder
• Llama - Providing basic language model architecture
• Vocos - Providing high-quality vocoder
• GRPO-Zero - Reinforcement learning algorithm implementation inspiration

If you use GLM-TTS in your research, please cite:

@misc{glmtts2025,
  title={GLM-TTS: Controllable & Emotion-Expressive Zero-shot TTS with Multi-Reward Reinforcement Learning},
  author={CogAudio Group Members},
  year={2025},
  publisher={Zhipu AI Inc}
}