OpenAI Speech-to-Text Models Launch with 25% Price Cut and Lower Error Rates

New GPT Transcribe models offer 25% lower costs and improved accuracy for batch and live audio, achieving a 3.31% word error rate.

Woofun AI reports that OpenAI has launched two new speech-to-text models, GPT Transcribe and GPT Live Transcribe, designed for batch processing and real-time captioning respectively. These models integrate audio topics and language cues to enhance recognition of technical terms, numbers, and accents in noisy environments.

Data from Artificial Analysis indicates GPT Transcribe achieves a 3.31% word error rate, a 0.7 percentage point improvement over the previous GPT-4o Transcribe. The pricing has been reduced by 25%, with costs set at $4.5 per 1000 minutes of audio.