Text-to-Speech

SenseAudio Text-to-Speech (TTS) API for converting text to natural speech. Supports synchronous and SSE streaming modes, multiple voices, emotion control, sp...

SenseAudio Text-to-Speech (TTS)

SenseAudio TTS converts text to natural, emotionally rich speech using a large language model. Supports 10+ emotions, streaming output (SSE), and fine-grained voice control.

Endpoint: POST https://api.senseaudio.cn/v1/t2a_v2 Auth: Authorization: Bearer $SENSEAUDIO_API_KEY Max text length: 10,000 characters

---

Request Parameters

Headers

Header	Required	Value
Authorization	yes	Bearer YOUR_API_KEY
Content-Type	yes	application/json

Body

Parameter	Type	Required	Description
model	string	yes	SenseAudio-TTS-1.0
text	string	yes	Text to synthesize. Supports <break time=500> pause tags
stream	boolean	yes	false = sync response; true = SSE streaming
voice_setting	object	yes	Voice configuration (see below)
audio_setting	object	no	Audio format configuration (see below)
dictionary	array	no	Polyphonic character corrections (cloned voices + TTS-1.5 only)

voice_setting

Parameter	Type	Default	Range	Description
voice_id	string	-	-	Voice ID (system or cloned)
speed	float	1.0	[0.5, 2.0]	Speech speed
vol	float	1.0	[0, 10]	Volume
pitch	int	0	[-12, 12]	Pitch adjustment
latex_read	boolean	false	-	Read LaTeX/MathML formulas aloud

audio_setting

Parameter	Type	Default	Options
format	string	mp3	mp3, wav, pcm, flac
sample_rate	int	32000	8000, 16000, 22050, 24000, 32000, 44100
bitrate	int	128000	32000, 64000, 128000, 256000 (MP3 only)
channel	int	2	1 (mono), 2 (stereo)

<break> Pause Tag

Insert pauses in text:

你好<break time=500>欢迎使用我们的服务

• time unit: milliseconds, min 100ms

---

Non-Streaming Response

{
  "data": {
    "audio": "hex-encoded audio data...",
    "status": 2
  },
  "extra_info": {
    "audio_length": 3500,
    "audio_sample_rate": 32000,
    "audio_size": 56000,
    "bitrate": 128000,
    "audio_format": "mp3",
    "audio_channel": 1,
    "word_count": 24,
    "usage_characters": 30
  },
  "base_resp": {"status_code": 0, "status_msg": "success"}
}

data.audio is hex-encoded. Decode: bytes.fromhex(audio_hex)

---

SSE Streaming Response

Each chunk: data: {"data":{"audio":"hex...","status":1},...}

Final chunk has status: 2 and includes extra_info.

---

Code Examples

curl (non-streaming)

curl -X POST https://api.senseaudio.cn/v1/t2a_v2 \
  -H "Authorization: Bearer $SENSEAUDIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "SenseAudio-TTS-1.0",
    "text": "道可道，非常道。名可名，非常名。",
    "stream": false,
    "voice_setting": {"voice_id": "male_0004_a"}
  }' -o response.json

jq -r '.data.audio' response.json | xxd -r -p > output.mp3

Python (non-streaming)

import requests

resp = requests.post(
    "https://api.senseaudio.cn/v1/t2a_v2",
    headers={"Authorization": "Bearer YOUR_API_KEY", "Content-Type": "application/json"},
    json={
        "model": "SenseAudio-TTS-1.0",
        "text": "道可道，非常道。",
        "stream": False,
        "voice_setting": {"voice_id": "male_0004_a"}
    }
)
result = resp.json()
audio_bytes = bytes.fromhex(result["data"]["audio"])
with open("output.mp3", "wb") as f:
    f.write(audio_bytes)

Python (SSE streaming)

import requests, json

with requests.post(
    "https://api.senseaudio.cn/v1/t2a_v2",
    headers={"Authorization": "Bearer YOUR_API_KEY", "Content-Type": "application/json"},
    json={"model": "SenseAudio-TTS-1.0", "text": "这是流式输出示例。", "stream": True,
          "voice_setting": {"voice_id": "male_0004_a"}},
    stream=True
) as r:
    with open("output.mp3", "wb") as f:
        for line in r.iter_lines():
            if line:
                line_str = line.decode("utf-8")
                if line_str.startswith("data: "):
                    chunk = json.loads(line_str[6:])
                    if chunk.get("data", {}).get("audio"):
                        f.write(bytes.fromhex(chunk["data"]["audio"]))

---

---

SenseAudio 文本转语音（TTS）

SenseAudio TTS 基于千亿参数大模型，将文字转化为自然流畅、情感丰富的语音。支持 10+ 种情感、流式输出（SSE）及精细化语音控制。

接口地址： POST https://api.senseaudio.cn/v1/t2a_v2 鉴权： Authorization: Bearer $SENSEAUDIO_API_KEY 最大文本长度： 10,000 字符

---

请求参数

请求头

参数名	必填	说明
Authorization	是	Bearer YOUR_API_KEY
Content-Type	是	application/json

请求体

参数名	类型	必填	说明
model	string	是	SenseAudio-TTS-1.0
text	string	是	待合成文本，支持 <break time=500> 停顿符
stream	boolean	是	false 同步；true SSE 流式
voice_setting	object	是	音色设置（见下表）
audio_setting	object	否	音频格式设置（见下表）
dictionary	array	否	多音字纠正（仅克隆音色 + TTS-1.5）

voice_setting（音色设置）

参数名	类型	默认值	范围	说明
voice_id	string	-	-	音色 ID（系统音色或克隆音色）
speed	float	1.0	[0.5, 2.0]	语速
vol	float	1.0	[0, 10]	音量
pitch	int	0	[-12, 12]	音调
latex_read	boolean	false	-	数学公式朗读

audio_setting（音频设置）

参数名	类型	默认值	选项
format	string	mp3	mp3, wav, pcm, flac
sample_rate	int	32000	8000/16000/22050/24000/32000/44100
bitrate	int	128000	32000/64000/128000/256000（仅 MP3）
channel	int	2	1（单声道）, 2（双声道）

<break> 停顿符

在文本中插入停顿：

你好<break time=500>欢迎使用我们的服务

• time 单位为毫秒，最小值 100ms

---

非流式响应

{
  "data": {"audio": "hex编码音频...", "status": 2},
  "extra_info": {
    "audio_length": 3500,
    "audio_sample_rate": 32000,
    "audio_size": 56000,
    "audio_format": "mp3",
    "word_count": 24,
    "usage_characters": 30
  },
  "base_resp": {"status_code": 0, "status_msg": "success"}
}

data.audio 为 hex 编码，解码：bytes.fromhex(audio_hex)

---

SSE 流式响应

每个数据块：data: {"data":{"audio":"hex...","status":1},...}

最后一个 chunk status: 2，包含完整 extra_info。

---

代码示例

curl（非流式）

curl -X POST https://api.senseaudio.cn/v1/t2a_v2 \
  -H "Authorization: Bearer $SENSEAUDIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "SenseAudio-TTS-1.0",
    "text": "道可道，非常道。名可名，非常名。",
    "stream": false,
    "voice_setting": {"voice_id": "male_0004_a"}
  }' -o response.json

jq -r '.data.audio' response.json | xxd -r -p > output.mp3

Python（非流式）

import requests

resp = requests.post(
    "https://api.senseaudio.cn/v1/t2a_v2",
    headers={"Authorization": "Bearer YOUR_API_KEY", "Content-Type": "application/json"},
    json={
        "model": "SenseAudio-TTS-1.0",
        "text": "道可道，非常道。",
        "stream": False,
        "voice_setting": {"voice_id": "male_0004_a"}
    }
)
audio_bytes = bytes.fromhex(resp.json()["data"]["audio"])
open("output.mp3", "wb").write(audio_bytes)