GPT-4o Mini Transcribe

GPT-4o Mini Transcribe is a lightweight speech-to-text model from OpenAI, built on the GPT-4o Mini architecture. It offers improved word error rates and better language recognition compared to the original Whisper models, with lower latency and cost suited for high-throughput transcription workflows. The model supports real-time audio streaming via WebSocket connections and accepts contextual prompts to improve transcription accuracy for domain-specific terminology.