Amphion 开放平台 API 文档Amphion Platform API Documentation

Amphion 开放平台提供高质量的语音 AI 服务，包括 TTS（文本转语音）和 SED（语音情感检测）。所有接口通过 RESTful API 提供，使用 API Key 进行身份验证。Amphion Platform provides high-quality voice AI services, including TTS (Text-to-Speech) and SED (Speech Emotion Detection). All endpoints are RESTful APIs authenticated via API Key.

认证方式Authentication

所有 API 请求需携带有效的 API Key。支持以下两种方式：All API requests require a valid API Key. Two methods are supported:

X-API-Key: sk-your-api-key-here

Authorization: Bearer sk-your-api-key-here

你可以在 控制台 → API Keys 页面创建和管理你的密钥。You can create and manage your keys in Console → API Keys.

Base URL

https://api.amphion.example.com

POST TTS — 文本转语音TTS — Text-to-Speech

将文本转换为高质量语音，基于 Qwen3-TTS 模型。Convert text to high-quality speech using Qwen3-TTS model.

端点Endpoint

POST /tts/v1/audio/speech

请求参数Request Parameters

响应Response

成功时返回音频二进制流，Content-Type 为 audio/mpeg 或 audio/wav。On success, returns audio binary stream with Content-Type audio/mpeg or audio/wav.

代码示例Code Examples

curl -X POST https://api.amphion.example.com/tts/v1/audio/speech \
  -H "X-API-Key: sk-your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen3-TTS",
    "input": "你好，欢迎使用 Amphion 开放平台。",
    "voice": "default"
  }' \
  --output output.mp3

import requests

resp = requests.post(
    "https://api.amphion.example.com/tts/v1/audio/speech",
    headers={"X-API-Key": "sk-your-api-key"},
    json={
        "model": "Qwen3-TTS",
        "input": "你好，欢迎使用 Amphion 开放平台。",
        "voice": "default",
    },
)

with open("output.mp3", "wb") as f:
    f.write(resp.content)

print(f"Status: {resp.status_code}, Size: {len(resp.content)} bytes")

const fs = require("fs");

const resp = await fetch("https://api.amphion.example.com/tts/v1/audio/speech", {
  method: "POST",
  headers: {
    "X-API-Key": "sk-your-api-key",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "Qwen3-TTS",
    input: "你好，欢迎使用 Amphion 开放平台。",
    voice: "default",
  }),
});

const buf = Buffer.from(await resp.arrayBuffer());
fs.writeFileSync("output.mp3", buf);
console.log(`Status: ${resp.status}, Size: ${buf.length} bytes`);

POST SED — 语音情感检测SED — Speech Emotion Detection

分析音频中的情感成分，基于 Zipformer 语音情感检测模型。Analyze emotional components in audio using Zipformer Speech Emotion Detection model.

端点Endpoint

POST /sed/v1/audio/analyze

请求参数Request Parameters

以 multipart/form-data 上传音频文件。Upload audio file as multipart/form-data.

响应示例Response Example

{
  "emotions": [
    { "label": "happy",   "score": 0.82 },
    { "label": "neutral", "score": 0.12 },
    { "label": "sad",     "score": 0.04 },
    { "label": "angry",   "score": 0.02 }
  ],
  "duration_seconds": 3.45,
  "model": "Zipformer-SED"
}

代码示例Code Examples

curl -X POST https://api.amphion.example.com/sed/v1/audio/analyze \
  -H "X-API-Key: sk-your-api-key" \
  -F "file=@audio.wav"

import requests

with open("audio.wav", "rb") as f:
    resp = requests.post(
        "https://api.amphion.example.com/sed/v1/audio/analyze",
        headers={"X-API-Key": "sk-your-api-key"},
        files={"file": ("audio.wav", f, "audio/wav")},
    )

data = resp.json()
for emo in data["emotions"]:
    print(f"{emo['label']}: {emo['score']:.2%}")

const fs = require("fs");
const FormData = require("form-data");

const form = new FormData();
form.append("file", fs.createReadStream("audio.wav"));

const resp = await fetch("https://api.amphion.example.com/sed/v1/audio/analyze", {
  method: "POST",
  headers: { "X-API-Key": "sk-your-api-key", ...form.getHeaders() },
  body: form,
});

const data = await resp.json();
data.emotions.forEach(e => console.log(`${e.label}: ${(e.score * 100).toFixed(1)}%`));

错误码说明Error Codes

当请求失败时，API 返回 JSON 格式的错误响应。When a request fails, the API returns a JSON error response.

{
  "detail": "Missing API key. Provide X-API-Key header or Authorization: Bearer <key>."
}

计费说明Pricing

API 调用按请求数计费，每次成功调用从账户余额中扣除相应费用。API calls are billed per request. Each successful call deducts from your balance.