ModelInfo
简体中文
全部模型

Amazon Nova 2 Sonic

Amazon 的语音到语音基础模型,用于构建自然的实时语音对话应用。

Amazonnova-2-sonic2025-12-02报告数据错误

规格

上下文
1M
最大输出
65.5K
输入价
$3/ 百万 Token
输出价
$12/ 百万 Token
发布日期
2025-12-02

上下文

上下文
1M
最大输入
—
最大输出
65.5K

价格 · USD / 百万 Token

输入
$3
输出
$12
缓存读取
—
缓存写入
—

输入输出

输入
文本音频
输出
文本音频

API

接口类型
realtime

思考

官方未公布

源信息

状态
可用
发布日期
2025-12-02
知识截止
—

能力

已确认
streaming
来源 · 官方文档核验 2026-10-10

调用方式

1 个 Provider

amazon-bedrockmodel = amazon.nova-2-sonic-v1:0

realtime
POSThttps://bedrock-runtime.us-east-1.amazonaws.com/model/{modelId}/invoke-with-bidirectional-stream

JSON

标准格式
nova-2-sonic.json
{
  "id": "nova-2-sonic",
  "object": "model",
  "created": 1764633600,
  "owned_by": "amazon",
  "name": "Amazon Nova 2 Sonic",
  "api": {
    "types": [
      "realtime"
    ]
  },
  "limits": {
    "context": 1000000,
    "input": null,
    "output": 65536
  },
  "modalities": {
    "input": [
      "text",
      "audio"
    ],
    "output": [
      "text",
      "audio"
    ]
  },
  "reasoning": {
    "supported": null,
    "efforts": []
  },
  "pricing": {
    "currency": "USD",
    "unit": "1M_tokens",
    "input": 3,
    "output": 12,
    "cache_read": null,
    "cache_write": null
  },
  "features": [
    "streaming"
  ],
  "info": {
    "status": "active",
    "release_date": "2025-12-02",
    "knowledge_cutoff": null,
    "description": "Amazon's speech-to-speech foundation model for building natural, real-time voice conversation applications.",
    "docs": "https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-sonic.html",
    "verified_at": "2026-10-10"
  }
}

官方来源

3