GPT-Realtime-2
Mis à jour le 2025-01-03Confiance : high
gpt-realtime-2openaivoice-modelsgpt-5-class-reasoningrealtime-audiowebrtc-apiknowledge-cutoffseptember-2024conversational-aidocument-contextchatgpt-iphonebrowser-tools
openai's advanced voice model described as their "first voice model with GPT-5-class reasoning," released for the webrtc-api in May 2026. Represents significant upgrade over previous real-time audio models with enhanced reasoning capabilities and document context support.
Key Characteristics
- Advanced Reasoning: Promoted as having "GPT-5-class reasoning" capabilities
- Knowledge Cutoff: September 30, 2024
- Real-time Audio: Optimized for low-latency voice interactions
- Document Context: Supports integration with document content for conversational exploration
Availability
Initially available only through the webrtc-api, not yet integrated into consumer applications like the ChatGPT iPhone app as of June 2026. This limited availability has driven development of independent playground-development tools like simon-willison's openai-webrtc-playground.
Applications
- document-context-audio conversations
- Real-time voice interactions with enhanced reasoning
- Browser-based AI assistants
- Experimental voice AI applications