GLM-4-Voice
by Zhipu AI
· GLM 4 family · released Oct 14, 2024
Open weights
Active
Free voice-in voice-out conversational model.
This model has no first-party API. Open-weight release: self-host from published weights.
Specifications
| Model type | Multimodal |
|---|---|
| Architecture | Dense |
| Parameters | 9 |
| Context window | 8.2K tokens |
| Max output | Not disclosed tokens |
| Knowledge cutoff | Not disclosed |
| License | ChatGLM License |
| Modalities | Text input Audio input Text output Audio output |
| Tags | - |
| End of life | Not disclosed |
| Last verified | Aug 20, 2026 |