Qwen-Audio 3.0 TTS Plus API

POST /v1/tasks

All models are called through the Unified Async API POST /v1/tasks endpoint; only the input fields differ (see input parameters below).

Model summary

Status	Online
HiAPI model ID	qwen-audio-3.0-tts-plus
Official Qwen model ID	qwen-audio-3.0-tts-plus
Launch price	$0.040 / 1,000 characters
Type	Audio generation (text-to-speech)

Qwen-Audio 3.0 TTS Plus targets high-quality speech production for voiceovers, video narration, and spoken content. HiAPI exposes complete speech controls and bills by effective character count.

Production guidance

Public parameter notes

The public contract on this page has passed live-task validation in HiAPI production.
Public input supports text, voice, format, sample_rate, volume, rate, pitch, seed, bit_rate, language_hints, and instruction.
format supports pcm, wav, mp3, and opus with mp3 as default; sample_rate defaults to 22050 Hz; bit_rate applies only to Opus.
Official Plus system voices: longanlingxin, longanlufeng.
Price: $0.040 / 1,000 characters; Alibaba counts each CJK ideograph as two billable characters and every other character as one.

Service acceptance results

Plus and Flash both passed full and minimal production requests.
MP3 and Opus artifacts downloaded successfully, and the showcase samples were copied to durable storage.
Effective-character charges matched the public price for every live task in this validation.

API boundaries

Provider credentials, internal request parameters, request IDs, and timestamps are not exposed.
Required internal metadata is handled server-side; SSML, cloned voices, and real-time streaming input are not public.
hiAPI currently accepts up to 20,000 Unicode code points for every format; PCM/WAV has no separate undocumented 600-character cap.

Best suited for

Professional voiceover

Suitable for high-quality speech generation for ads, product demos, and brand content.

textvoiceinstruction

Video narration

Suitable for explainers, tutorials, and short-video narration.

textvoiceratepitch

Spoken content

Suitable for spoken articles, podcast segments, and content submitted in sections.

textvoice

Request parameters

POST https://api.hiapi.ai/v1/tasks

# Create an async task — returns taskId immediately (generation runs async)
# Replace YOUR_API_KEY with your account API key
curl -X POST "https://api.hiapi.ai/v1/tasks" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "qwen-audio-3.0-tts-plus",
  "input": {
    "text": "欢迎来到 HiAPI 音频模型实验室。我们正在比较两种语音合成模型在中文表达、英文切换和数字读法上的表现。Today, we test clarity, rhythm, and natural pauses. 版本三点零，样本编号二零二六。",
    "voice": "longanlingxin",
    "format": "mp3",
    "sample_rate": 24000,
    "volume": 50,
    "rate": 1,
    "pitch": 1,
    "seed": 2026,
    "language_hints": [
      "zh"
    ],
    "instruction": "请用沉稳、清晰、有亲和力的产品解说语气朗读。中文表达自然，英文发音准确，在句号处稍作停顿。"
  }
}'

import requests, json

# Business params — fields vary by model (see the model page)
payload = json.loads(r'''
{
  "model": "qwen-audio-3.0-tts-plus",
  "input": {
    "text": "欢迎来到 HiAPI 音频模型实验室。我们正在比较两种语音合成模型在中文表达、英文切换和数字读法上的表现。Today, we test clarity, rhythm, and natural pauses. 版本三点零，样本编号二零二六。",
    "voice": "longanlingxin",
    "format": "mp3",
    "sample_rate": 24000,
    "volume": 50,
    "rate": 1,
    "pitch": 1,
    "seed": 2026,
    "language_hints": [
      "zh"
    ],
    "instruction": "请用沉稳、清晰、有亲和力的产品解说语气朗读。中文表达自然，英文发音准确，在句号处稍作停顿。"
  }
}
''')

# Create an async task — returns taskId immediately (generation runs async)
resp = requests.post(
    "https://api.hiapi.ai/v1/tasks",
    headers={"Authorization": "Bearer YOUR_API_KEY", "Content-Type": "application/json"},
    json=payload,
)
print(resp.json())  # Returns { code, message, data: { taskId } }

// Create an async task — returns taskId immediately (generation runs async)
// Replace YOUR_API_KEY with your account API key
const resp = await fetch("https://api.hiapi.ai/v1/tasks", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
  "model": "qwen-audio-3.0-tts-plus",
  "input": {
    "text": "欢迎来到 HiAPI 音频模型实验室。我们正在比较两种语音合成模型在中文表达、英文切换和数字读法上的表现。Today, we test clarity, rhythm, and natural pauses. 版本三点零，样本编号二零二六。",
    "voice": "longanlingxin",
    "format": "mp3",
    "sample_rate": 24000,
    "volume": 50,
    "rate": 1,
    "pitch": 1,
    "seed": 2026,
    "language_hints": [
      "zh"
    ],
    "instruction": "请用沉稳、清晰、有亲和力的产品解说语气朗读。中文表达自然，英文发音准确，在句号处稍作停顿。"
  }
}),
});
console.log(await resp.json()); // Returns { code, message, data: { taskId } }

model string required

Fixed value: qwen-audio-3.0-tts-plus.

example qwen-audio-3.0-tts-plus

input object required

Speech synthesis parameters for Qwen-Audio 3.0 TTS Plus.

text string required

Non-empty text to synthesize. hiAPI accepts up to 20,000 Unicode code points per task for every output format. Price: $0.040 / 1,000 characters.

voice enum required

Official system voice supported by the Plus version. View available voices

default longanlingxin enum: longanlingxinlonganlufeng

format enum optional

Output format: PCM, WAV, MP3, or Opus.

default mp3 enum: pcmwavmp3opus

sample_rate enum optional

Output sample rate in Hz.

default 22050 enum: 80001600022050240004410048000

volume integer optional

Volume from 0 to 100.

default 50

rate number optional

Speech-rate multiplier from 0.5 to 2.0.

default 1

pitch number optional

Pitch multiplier from 0.5 to 2.0.

default 1

seed integer optional

Synthesis seed from 0 to 65,535.

default 0

bit_rate integer optional

Opus bit rate from 6 to 510 kbps; only applies when format=opus.

default 32

language_hints string[] optional

Optional language hint with one official language code; Qwen processes only the first array item.

enum: zhenfrdejakoruptthid +6

instruction string optional

Optional control for dialect, emotion, or speaking style.

callback object optional

Optional callback configuration for terminal task notifications.

url string required

HTTPS endpoint required when callback is set.

example https://your-domain.com/hiapi/callback

when enum optional

Callback timing; use final.

default final enum: final

Plus available voices

Official system voices for this tier.

voice ID	Status
longanlingxin	`Online`
longanlufeng	`Online`

Example requests

MP3 speech request

Uses default MP3 output at 22.05 kHz.

Request body

{
  "model": "qwen-audio-3.0-tts-plus",
  "input": {
    "text": "欢迎来到 HiAPI 音频模型实验室。我们正在比较两种语音合成模型在中文表达、英文切换和数字读法上的表现。Today, we test clarity, rhythm, and natural pauses. 版本三点零，样本编号二零二六。",
    "voice": "longanlingxin",
    "format": "mp3",
    "sample_rate": 24000,
    "volume": 50,
    "rate": 1,
    "pitch": 1,
    "seed": 2026,
    "language_hints": [
      "zh"
    ],
    "instruction": "请用沉稳、清晰、有亲和力的产品解说语气朗读。中文表达自然，英文发音准确，在句号处稍作停顿。"
  }
}

PCM request with speech controls

Sets speech rate, pitch, and volume explicitly.

Request body

{
  "model": "qwen-audio-3.0-tts-plus",
  "input": {
    "text": "This is a speech-control test.",
    "voice": "longanlingxin",
    "format": "pcm",
    "sample_rate": 24000,
    "volume": 50,
    "rate": 1.1,
    "pitch": 1,
    "seed": 0
  }
}

Getting the result

A successful submission returns taskId immediately without waiting for audio generation.
Prefer callback.url for terminal notifications in production; use GET /v1/tasks/:id for local polling.
After status=success, download the audio from output[].url and copy it to storage when needed.

FAQ

How do I call this model?

Create an API key in the dashboard, then send the model ID and input to POST /v1/tasks. Use the returned taskId to check status, or configure a callback for the terminal notification.

What is the launch price?

The Plus price is $0.040 / 1,000 characters, billed by effective character count.

How are characters counted?

Each CJK ideograph counts as two characters; letters, digits, punctuation, spaces, kana, and other characters count as one.

Which voices are supported?

Official Plus system voices are longanlingxin, longanlufeng. View available voices

What is the maximum text length?

hiAPI accepts up to 20,000 Unicode code points per task for PCM, WAV, MP3, and Opus.

Are real-time streaming or SSML supported?

They are outside the public launch contract. HiAPI exposes only the unified asynchronous task API.

Are failed tasks charged?

Tasks that reach a failed terminal state are not billed. If credits were reserved at submission, they are returned automatically after failure.