Speech to text

​

Endpoint

​

Request parameters

• model string required
  Model name, fixed to step-asr.
• response_format string required
  Output format. Supported: json, text, srt, vtt.
• file File required
  Audio file. Supported formats: flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, webm, aac, opus.
  File size must be under 100 MB.
• hotwords string optional
  Hotword list as a JSON string that parses to an array, e.g., ["1","2","3","4","abc"].

​

Response

{ "text": "Test recording" }

Test recording

1
00:00:00,000 --> 00:00:03,240
Test recording

WEBVTT

00:00:00,000 --> 00:00:03,240
Test recording

​

Example

curl -L 'https://api.stepfun.ai/v1/audio/transcriptions' \
-H "Authorization: Bearer $STEP_API_KEY" \
-F 'model="step-asr"' \
-F 'response_format="json"' \
-F 'file=@"sample.mp3"'