openai-whisper-api

Transcribe audio files into text by calling the OpenAI Whisper API via curl. Execute the transcribe.sh script with your audio file to generate a plain text or JSON transcript for meetings, podcasts, or voice notes.

The Whisper API transcription supports m4a, wav, and mp3 audio formats. You can pass any of these file types to the curl-based workflow to convert speech into searchable text transcripts.

Yes, converting speech to text requires an OPENAI_API_KEY. You must provide this key in the curl command environment to authenticate requests to the OpenAI Whisper API endpoint.

Yes, you can improve transcription accuracy by providing optional language hints or prompts. These parameters guide the Whisper API model to better recognize specific speech patterns during audio-to-text conversion.

Automate speech-to-text transcription by integrating the transcribe.sh curl script into your data processing pipeline. The script handles model selection and outputs plain text or JSON, enabling batch processing of audio content.

The transcription workflow outputs plain text by default but also supports JSON format. You can specify the desired output format when calling the Whisper API via curl to suit your downstream data processing needs.