Ask HN: Is there a duplex voice mode in Claude Code?

An Ask HN thread titled “Is there a duplex voice mode in Claude Code?” is live on Hacker News at item 49178894. At capture it sat at 1 point with 0 comments, so the question is still unanswered in that thread. The item is tagged against the Claude, Codex, and Fable cluster of AI coding tooling discussion.

Duplex voice mode, in product terms, means full-duplex audio: the system can listen and speak at the same time instead of strict turn-taking (half-duplex). For a CLI agent like Claude Code, that would imply continuous speech input, barge-in (interrupt while the agent is talking), and low-latency spoken replies wired into the same session that already runs tools, edits, and shell commands—not a separate voice chat bolted on later.

For engineers, the practical ask is whether voice can sit on the critical path of coding work: dictate intent while watching diffs, interrupt a long agent monologue, and keep hands on the keyboard or off it without losing session state. If duplex is missing, voice remains a novelty; if it exists or is coming, it changes how people drive long agent loops and review tool output without context-switching to typed prompts.

The competitive frame is other AI coding products in the same HN cluster—Claude-side tooling, Codex-class coding agents, and adjacent stacks people group under Fable—where voice is increasingly a table-stakes interface claim even when the underlying agent loop is still text-first. A zero-comment Ask HN with a single point does not prove a gap or a feature; it only shows demand is being probed in public before anyone has confirmed product reality.

What to watch next is whether the thread gets a first-party or reliable second-hand answer: supported natively, experimental flag, third-party wrapper, or not available. Until that lands, treat duplex voice in Claude Code as an open product question, not a shipped capability, and validate any claim against official docs or a live demo rather than the Ask HN title alone.