GLM-4-Voice

首个端到端语音模型，可理解和生成中英文语音，用于实时对话。

Zhipu AIglm-4-voice报告数据错误

规格

上下文: 8.2K
最大输出: 4.1K
输入价: ¥80/ 百万 Token
输出价: ¥80/ 百万 Token
发布日期: —

上下文

上下文: 8.2K
最大输入: —
最大输出: 4.1K

价格 · CNY / 百万 Token

输入: ¥80
输出: ¥80
缓存读取: —
缓存写入: —

输入输出

输入: 音频文本
输出: 音频

API

接口类型: audio

思考

官方未公布

源信息

状态: 可用
发布日期: —
知识截止: —

能力

官方未公布

调用方式

zhipumodel = glm-4-voice

audio

POSThttps://open.bigmodel.cn/api/paas/v4/chat/completions

JSON

glm-4-voice.json

{
  "id": "glm-4-voice",
  "object": "model",
  "created": null,
  "owned_by": "zhipu",
  "name": "GLM-4-Voice",
  "api": {
    "types": [
      "audio"
    ]
  },
  "limits": {
    "context": 8192,
    "input": null,
    "output": 4096
  },
  "modalities": {
    "input": [
      "audio",
      "text"
    ],
    "output": [
      "audio"
    ]
  },
  "reasoning": {
    "supported": null,
    "efforts": []
  },
  "pricing": {
    "currency": "CNY",
    "unit": "1M_tokens",
    "input": 80,
    "output": 80,
    "cache_read": null,
    "cache_write": null
  },
  "features": [],
  "info": {
    "status": "active",
    "release_date": null,
    "knowledge_cutoff": null,
    "description": "First end-to-end speech model that understands and generates Chinese and English speech for real-time conversation.",
    "docs": "https://docs.bigmodel.cn/cn/guide/models/sound-and-video/glm-4-voice",
    "verified_at": "2026-10-11"
  }
}

官方来源