Codex's new voice mode picks up sniffing, tongue clicks, coughs and gulps in real time
altryne · x · 2026-09-05
Developer altryne tested OpenAI Codex's new voice mode and was stunned by its environmental audio recognition:
- The mode detects non-verbal sounds in real time — sniffing, tongue clicks, coughs, gulps — not just speech, which he calls "crazy witchcraft."
- It is built into the Codex coding agent workflow, with the author apparently driving tasks via voice.
- A demo video accompanies the post.
The detail suggests OpenAI is bringing real-time multimodal voice into its coding agent, with far finer non-speech audio granularity than typical voice assistants.
More from coding & agent
- The Story of VS Code documentary debuts, noting Antigravity, Cursor, Windsurf are all forks — 0xkarasy · 2026-09-05
- TRACES agent benchmark grades live execution loops, not answers — SucceededMind · 2026-09-05
- Dev says GPT-6 Astra found 4.5x to 176x speedups in his codebase in 5 minutes — charliermarsh · 2026-09-05
- Update Codex CLI for remote connections to access Astra — benjamin_warner · 2026-09-05
- OpenAI Announces GPT-6 Astra Hackathons in SF and NYC, $50K Credits Top Prize — OpenAIDevs · 2026-09-05
- Astra implements entire design system in Figma, taking over designers' grunt work — AIandDesign · 2026-09-05