TTS 快速开始

本指南将使用 Shisa TTS 生成您的第一段语音音频。您需要一个 API 密钥 — 在 Shisa 平台创建。新账户包含 $10 的免费额度。

1. 设置您的端点和密钥

使用标准 Bearer 令牌对每个请求进行身份验证：

Endpoint:  https://api.shisa.ai/tts
Auth:      Authorization: Bearer YOUR_API_KEY

提示

请勿将 API 密钥纳入源代码管理。在实际应用程序中，应从环境变量（例如 SHISA_API_KEY）中读取。

2. 生成语音

发送包含您所需 voice_id、text 和 format 的 POST 请求。以下示例使用声音 61ba1141-60aa-4bc3-a3b3-be1ec20700b3。

curl

# List available voices
curl -s -X GET "https://api.shisa.ai/tts/voices" \
  -H "Authorization: Bearer YOUR_API_KEY" | jq .

# Generate speech
curl -s -X POST "https://api.shisa.ai/tts" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "voice_id": "61ba1141-60aa-4bc3-a3b3-be1ec20700b3",
    "format": "mp3",
    "stream": false,
    "text": "こんにちは。Shisa APIへようこそ。"
  }' \
  --output speech.mp3

# Stream audio directly to a player
curl -s -X POST "https://api.shisa.ai/tts" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "voice_id": "61ba1141-60aa-4bc3-a3b3-be1ec20700b3",
    "format": "mp3",
    "stream": true,
    "text": "ストリーミングテストです。"
  }' \
  --output - | ffplay -nodisp -autoexit -

Python

import requests

API_URL = "https://api.shisa.ai"
API_KEY = "YOUR_API_KEY"
HEADERS = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json"
}

# List available voices
def list_voices():
    response = requests.get(f"{API_URL}/tts/voices", headers=HEADERS)
    return response.json()

# Generate speech
def generate_speech(text, voice_id="61ba1141-60aa-4bc3-a3b3-be1ec20700b3", format="mp3", stream=False):
    response = requests.post(
        f"{API_URL}/tts",
        headers=HEADERS,
        json={
            "voice_id": voice_id,
            "format": format,
            "stream": stream,
            "text": text
        },
        stream=stream
    )

    output_file = f"output.{format}"
    with open(output_file, "wb") as f:
        if stream:
            for chunk in response.iter_content():
                f.write(chunk)
        else:
            f.write(response.content)

    return output_file

# Example usage
voices = list_voices()
print(voices)

audio_file = generate_speech(
    "お客様の声を大切にしています。",
    voice_id="61ba1141-60aa-4bc3-a3b3-be1ec20700b3"
)

JavaScript

const API_URL = 'https://api.shisa.ai';
const API_KEY = 'YOUR_API_KEY';
const headers = {
  'Authorization': `Bearer ${API_KEY}`,
  'Content-Type': 'application/json',
};

// List available voices
const listVoices = async () => {
  const response = await fetch(`${API_URL}/tts/voices`, { headers });
  return response.json();
};

// Generate speech
const generateSpeech = async (text, voiceId = '61ba1141-60aa-4bc3-a3b3-be1ec20700b3', format = 'mp3') => {
  const response = await fetch(`${API_URL}/tts`, {
    method: 'POST',
    headers,
    body: JSON.stringify({
      voice_id: voiceId,
      format,
      stream: true,
      text,
    }),
  });

  // Handle streaming response
  const reader = response.body.getReader();
  const chunks = [];

  while (true) {
    const { done, value } = await reader.read();
    if (done) break;
    chunks.push(value);
  }

  // Combine chunks and create audio blob
  const blob = new Blob(chunks, { type: `audio/${format}` });
  const url = URL.createObjectURL(blob);

  // Play audio
  const audio = new Audio(url);
  audio.play();
};

// Example usage
const voices = await listVoices();
console.log(voices);

await generateSpeech('ようこそ、Shisa APIへ。', '61ba1141-60aa-4bc3-a3b3-be1ec20700b3');

3. 播放音频

响应是 二进制音频，而非 JSON — 请求会将其直接写入文件（speech.mp3）。使用任意音频工具播放：

# Play the generated audio
ffplay -nodisp -autoexit speech.mp3

# Or stream directly
curl -s -X POST "https://api.shisa.ai/tts" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"voice_id": "61ba1141-60aa-4bc3-a3b3-be1ec20700b3", "format": "mp3", "stream": true, "text": "ストリーミングテストです。"}' \
  --output - | ffplay -nodisp -autoexit -

备注

流式传输（"stream": true）仅适用于 streaming: true 的声音。请在声音目录中查看声音的功能。

后续步骤

在声音中浏览所有可用的声音。
在 API 参考中查看所有参数、响应格式和错误代码。
在身份验证中了解各服务通用的认证头约定。

1. ​设置​您​的​端点​和​密钥​

2. 生成语​音​

3.​ ​播放音频​

后续​步​骤​

1. 设置您的端点和密钥

2. 生成语音

3. 播放音频

后续步骤