mirror of
https://github.com/openai/codex.git
synced 2026-10-27 18:48:37 +00:00
Model realtime session output voices with an enum and map by parser so v1 uses fathom and v2 uses alloy. Co-authored-by: Codex <noreply@openai.com>