Full-duplex voice model for natural conversations that delegates reasoning and tools to a backend agent
Specification
openai listing
Context
0
Max output
0
Released
2026-09-10
Knowledge cutoff
2025-07-31
Retires
2027-09-10
Open weights
no
Input
text, audio
Output
text, audio
Available from 2 hosts
More from OpenAI
most-hosted first