Z.aiGLM Speech
GLM-TTS
A two-stage speech system with voice cloning and streaming inference.
Practical use
Practical guidance. Use phoneme controls for names and words with several pronunciations. Review short replies and long narration separately; natural delivery in one does not establish suitable timing in the other.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | glm-ttszai-org/GLM-TTS |
| Architecture | A language model generates speech tokens; flow matching produces the acoustic representation.GLM-TTS official documentation |
| Control | Hybrid phoneme and text input; Chinese and English mixed text.GLM-TTS official documentation |
| Reference | The card describes three-to-ten-second voice prompts.GLM-TTS official documentation |
| Licence | MIT.GLM-TTS official documentation |
Related models
- GLM-ASR-2512
Z.ai · GLM Speech
A hosted recognizer for multilingual audio transcription.
Launch date not confirmedAPI - GLM-ASR-Nano-2512
Z.ai · GLM Speech
A 1.5-billion-parameter downloadable speech recognizer.
Launch date not confirmedDownloadable