A speech-to-text model using GPT-4o for transcribing audio. It offers improved word error rate, better language recognition, and higher accuracy compared to the original Whisper models. Use it for more precise transcripts.