Talk to your code

The heart of Verbative: your voice drives the Claude Code CLI, and you stay in control of every step. Wake it with a word, dictate your prompt, hear Claude's replies and permission requests spoken back, and answer them out loud — no keyboard, no loss of visibility.

Verbative Hands-FreePut on your headphones and build

Two ways your prompt is transcribed

The short voice commands are always recognized on-device. How your dictated prompt is transcribed depends on the Transcription setting in the panel:

Local (on-device)— whisper.cpp transcribes your prompt on your Mac and types the text into the CLI. Nothing is sent to any cloud service. Available on every plan (the same daily command limit applies); it's the default on Advanced, where you can also choose which whisper model to use (Large v3 Turbo is the default). The panel shows a live transcript as you speak.
Cloud— taps Claude Code's own built-in voice dictation, which sends the prompt audio to Anthropic (exactly as if you used Claude Code's dictation by hand). The default on Free.

Your choice of whisper model is available on every plan — tiny for speed, Large v3 for accuracy and translation. See Configure to pick a model and set your audio devices.

Default Channel · Agent 1

Transcribing on-device…local

● Orange = still being transcribedCancelSend ↵

Speed 1.3×

Local mode — your prompt transcribed on-device, with a live transcript you can edit before it's sent.

Speak any language

Dictate in your native language and Verbative translates the prompt to English on-device before it ever reaches Claude — always on, on every plan, with nothing to switch on. The voice commands themselves stay English so nothing gets ambiguous.

Default Channel · Agent 1

Listening — any language…local

Deutsch detected→ English

● Orange = your words, any languageCancelSend ↵

Dictate in any language — Verbative translates it to English on-device before it reaches Claude.

What you can do by voice

Dictate & submit prompts— “listen” … “stop listening”.
Answer permission prompts— “approve” / “deny”, with a spoken “explain” if you want detail first.
Spoken status & summaries— “update” for a progress recap while Claude works; summaries are read aloud when it finishes.
Interrupt or cancel— “stop claude” interrupts a running task; “hold on” cancels a mis-fired listen without touching Claude.
Replay— “repeat” replays the last spoken cue; “repeat review” replays the last Advisory Board batch.
Re-hear a request— “open queues” re-speaks the permission request that's currently waiting, in case you missed it.
Control the voice— “skip”, “pause” and “resume” act on whatever cue is playing.
Editable commands Advanced— rename any trigger phrase to your own wording.

Voice commandson-device

Listen & dictate

listenWake — start a promptstop listeningSend the prompthold onCancel listening

Approve by voice

approveAllow the pending tooldenyBlock itexplainHear what it would do firstopen queuesRe-hear the open permission cue

Stay in flow

updateLive progress recapstop claudeInterrupt the running taskrepeatReplay the last cueskipSkip the cue playing nowpausePause the cue · resume to continue

Toggles

advisory boardSix-advisor design reviewproject roadmapLive spec → build → ship board

resumerepeat review+ more

Advanced: tap ✎ to rename any phrase to whatever feels natural

The full voice-command vocabulary — Advanced users can rename any trigger.

See voice control in action →