openai-whisper-api

To transcribe audio to text, run the provided transcription script with your local audio file path. The script sends a multipart form data request to the OpenAI Whisper API via curl, returning either plain text or JSON transcripts based on your configuration.

You need curl installed and a valid OPENAI_API_KEY environment variable. The script sends audio files as multipart form data to the api.openai.com endpoint to generate accurate text transcripts without manual typing.

Yes, the audio transcription process supports configurable model selection, language specification, and custom prompts. These options are passed via curl to the OpenAI API to improve transcription accuracy for your specific audio files.

The script supports outputting either plain text or JSON transcripts from the OpenAI Whisper API. You can configure the output format to integrate the resulting transcription into your workflows, chatbots, or media processing pipelines.

Yes, the transcription script is designed to process local audio files. You pass the local file path as an argument to the script, which then handles the multipart upload to the OpenAI API via curl.

Using curl to send multipart form data to the Whisper API provides a lightweight, dependency-free method for audio transcription. It enables quick integration into shell scripts and automated media processing pipelines without requiring heavy SDKs.