OpenAI releases GPT-Realtime-1.5 and GPT-Audio-1.5 models

AI Impact Summary

OpenAI has released the gpt-realtime-1.5 and gpt-audio-1.5 models, building upon previous versions with improvements in instruction following, multilingual support, and tool calling. These models are designed for voice-first applications and are accessible via the existing chat completion APIs in Microsoft Foundry, offering low-latency interactions. Developers should prioritize migrating to these new models to leverage the enhanced capabilities and maintain real-time performance for voice-based workflows.