OpenAI Voice API Major Upgrade: More Accurate Transcription, 40% Faster Agent Speed
OpenAI launched two API updates to enhance the performance of AI agents in voice interaction and complex tasks. The new real-time model gpt-realtime-1.5 and its accompanying audio model significantly improve the reliability of voice commands. Internal testing shows that the new model has improved digit and letter transcription accuracy by about 10%, logic audio task accuracy by 5%, and instruction execution accuracy by 7%.