HighCapability
OpenAI releases gpt-realtime-1.5 and gpt-audio-1.5 models for voice-first apps
Action Required
Developers using voice-based applications within Microsoft Foundry need to migrate to these new models to leverage improved performance and capabilities.
AI Impact Summary
OpenAI has released two new models, gpt-realtime-1.5 and gpt-audio-1.5, focused on voice-first applications. These models offer improvements in instruction following, multi-lingual support, and tool calling, particularly for real-time interactions. The release is through the Microsoft Foundry chat completion APIs, expanding capabilities for voice-based applications and enabling diarization for real-time conversations.
OpenAI releases gpt-realtime-1.5 and gpt-audio-1.5 models for voice-first apps
Affected Systems
gpt-realtime-1.5
- Date
- 23 Feb 2026
- Change type
- capability
- Severity
- high