OpenAI releases voice Agent

On March 21 at 01:00 (UTC+8), OpenAI conducted a technical live broadcast and released three new voice models specifically for developing AI Voice Agents. Two are voice-to-text models, GPT-40 Transcribe and GPT-4 Mini Transcribe; one is a text-to-speech model, GPT-40 Mini TTS. It's worth mentioning that developers can control the vocal emotion and style of the GPT-40 Mini TTS model. OpenAI has added a powerful streaming mode to its voice-to-text API, allowing developers to input continuous audio streams into the model in real time, and the model can also return continuous text and responses in real time. This feature of real-time interaction is very helpful for applications that require immediate feedback, such as real-time voice dialogue systems, transcription of voice meetings etc. (AIGC Open Community)