HighCapability
OpenAI releases GPT-Realtime-1.5 and GPT-Audio-1.5 models
Action Required
Developers can now build more responsive and natural voice-based applications with improved performance and capabilities.
AI Impact Summary
OpenAI has released two new models, gpt-realtime-1.5 and gpt-audio-1.5, focused on improving real-time interactions and multi-lingual support. These models are designed for voice-first applications and offer enhanced instruction following, tool calling, and reduced latency. Developers can leverage these models through the existing chat completion APIs in Microsoft Foundry, marking a significant advancement in real-time audio and speech capabilities.
OpenAI releases GPT-Realtime-1.5 and GPT-Audio-1.5 models
Affected Systems
GPT-Realtime-1.5
- Date
- 23 Feb 2026
- Change type
- capability
- Severity
- high