Skip to main content

​
语音合成 TTS

POST /v1/audio/speech
参数类型必填说明
modelstring是如 tts-1、tts-1-hd
inputstring是待合成的文本
voicestring是alloy、echo、fable、onyx、nova、shimmer
response_formatstring否mp3、opus、aac、flac
speednumber否0.25 ~ 4.0
curl https://your-domain.com/v1/audio/speech \
  -H "Authorization: Bearer sk-xxxx" \
  -H "Content-Type: application/json" \
  -d '{"model":"tts-1","input":"你好","voice":"alloy"}' \
  --output speech.mp3

​
语音识别 STT

POST /v1/audio/transcriptions 以 multipart/form-data 上传：
字段必填说明
file是音频文件
model是如 whisper-1
language否ISO-639-1
response_format否json、text、srt、vtt
curl https://your-domain.com/v1/audio/transcriptions \
  -H "Authorization: Bearer sk-xxxx" \
  -F file="@audio.mp3" \
  -F model="whisper-1"