Audio Transcription SRT

ARK Platform Example / Audio Transcription SRT

With the ARK API, you can generate transcriptions of audio files in various formats including SRT (SubRip Text), which is a popular subtitle format that includes timestamps.

Using the response_format parameter with the value "srt" enables you to get transcriptions with properly formatted timestamps that can be directly used for subtitling videos.

• Copy

  
  # NOTE: This code assumes an "audio.mp3" file exists in the current path with recorded speech.
  
  import openai
  
  ark_api_key = "API_KEY"
  ark_base_url = "https://api.ark-labs.cloud/api/v1"
  
  client = openai.OpenAI(api_key=ark_api_key, base_url=ark_base_url)
  
  with open("Rozmowa_1.mp3", "rb") as audio_file:
      print("Requesting transcription in SRT format...")
      response = client.audio.transcriptions.create(
          model="whisper-1",
          file=audio_file,
          language="pl",
          response_format="srt"  # Request SRT format
      )
  
  print("Transcription (SRT):")
  print(response)  # SRT is plain text, so directly printing the response