AI 音频

语音转文字

音频转写为文字(Whisper),返回文本、分段与语言

把一段音频异步转写为文字,返回全文、分段时间轴与识别语言。提交后用返回的 task_id 轮询。

Base URL:https://api.aiclonevoicefree.com | 鉴权:Authorization: Bearer sk_...

POST /api/v2/voice/transcribe

字段类型必填说明
audio_urlstring✅音频 URL(也接受别名 url);时长由服务端实测
languagestring⬜语言代码,留空自动检测

计费

按音频时长 1 积分/秒。提交时服务端从音频 URL 实测时长计费(cost = 向上取整(实测秒数)),无需自报时长;无法读取时长时提交返回 audio_probe_failed(请确认 URL 可公开访问)。提交时预扣并校验余额,不足返回 402;任务失败自动退款。

curl -X POST https://api.aiclonevoicefree.com/api/v2/voice/transcribe \
  -H "Authorization: Bearer sk_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{ "audio_url": "https://your-cdn.com/speech.mp3" }'

响应 202 → { "task_id": "...", "status": "pending", "capability": "voice", "action": "transcribe" }

取结果

轮询 GET /api/v2/voice/transcribe/{task_id}(注意:转写有专属状态端点,不走统一的 /tasks):

{
  "task_id": "...",
  "status": "completed",
  "capability": "voice",
  "action": "transcribe",
  "_type": "voice.transcribe",
  "text": "完整转写文本……",
  "language": "zh",
  "segments": [{ "start": 0.0, "end": 3.2, "text": "……" }]
}

On this page