Dictare / Docs / CLI Reference

CLI Reference¶

dictare agent¶

Launch an AI coding agent with voice input.

dictare agent <name> [options]

Option	Description
`--profile`, `--type`, `-t`	Agent profile override
`--verbose`, `-v`	Enable debug output
`--server`, `-s`	Engine OpenVIP URL
`--continue`, `-C`	Resume previous agent session
`--live-dangerously`	Skip agent permission prompts (YOLO mode)

dictare agent ozzy
dictare agent ozzy --continue
dictare agent gilmour --profile codex --verbose
dictare agent bowie -- claude --model opus

dictare service¶

Manage the Dictare background service.

dictare service <action>

Action	Description
`install`	Install the system service
`uninstall`	Remove the system service
`start`	Start the service
`stop`	Stop the service
`restart`	Restart the service and wait for engine readiness
`status`	Show service status
`logs`	View service logs

dictare service start
dictare service status
dictare service logs

dictare upgrade¶

Install a new Dictare runtime, activate it atomically, and repair service integration.

dictare upgrade [options]

Option	Description
`--version`	Install a specific Dictare version
`--from-path`	Install from a local wheel, sdist, or path
`--extras`	Comma-separated extras override, for example `mlx`
`--reinstall`	Reinstall into an existing version directory
`--no-restart`	Leave services stopped after activation

dictare upgrade
dictare upgrade --version 0.6.0rc1
dictare upgrade --reinstall --extras mlx

dictare rollback¶

Switch back to the previous runtime.

dictare rollback
dictare rollback --no-restart

dictare repair¶

Repair runtime-store shims, service files, and launcher integration.

dictare repair
dictare repair --no-start
dictare repair --force-bundle-migration

dictare uninstall¶

Move Dictare runtime and service state aside while preserving config and user data by default.

dictare uninstall
dictare uninstall --wipe-config

Use this before a clean reinstall. Homebrew users can then remove and reinstall the Homebrew entry point:

brew uninstall dictare
brew install dictare
dictare setup

dictare speak¶

Text-to-speech output.

dictare speak "Hello, world" [options]

Option	Description
`--list-engines`	List available TTS engines
`--list-voices`	List voices for the current engine
`--language`	Language code
`--speed`	Speech speed in words per minute
`--engine`	TTS engine override for this request
`--voice`	Speaker name
`--timeout`	Request timeout in seconds
`--quiet`	Suppress status messages
`--verbose`	Echo spoken text to stderr

dictare speak "Build complete"
dictare speak "Test passed" --engine kokoro
echo "Deploy complete" | dictare speak --engine say
dictare speak stop
dictare speak --list-engines
dictare speak --list-voices

dictare transcribe¶

dictare transcribe [options]

Option	Description
`--auto-submit`	Print the first transcription immediately instead of waiting for submit
`--url`, `-u`	OpenVIP server URL
`--agent-id`, `-a`	Custom agent ID
`--verbose`, `-v`	Echo output to stderr

dictare transcribe                # Transcription to stdout
dictare transcribe --auto-submit  # Print each transcription immediately

The Unix Pipe Dream¶

This one's for you, Simon.

dictare transcribe | llm | dictare speak

Speak into the mic. Your words flow through llm, and the answer comes back as speech. Voice in, voice out — three commands, zero code.

dictare status¶

Show engine health and connected agents.

dictare status [options]

Option	Description
`--json`	Output as JSON

dictare status
dictare status --json

dictare logs¶

View Dictare logs.

dictare logs [options]

Option	Description
`--follow`, `-f`	Follow log output (like `tail -f`)
`--last`	Show last N lines
`--name`	Filter by logger name
`--source`	Filter by source
`--raw`	Show raw log format
`--tts`	Show TTS-specific logs

dictare logs -f
dictare logs --last 50
dictare logs --name pipeline

dictare models¶

Manage STT and TTS models.

dictare models [action]

Lists available models, downloads new ones, or deletes cached models.

dictare models              # List available models
dictare models list         # List available models
dictare models pull <model> # Download a model
dictare models rm <model>   # Remove a cached model
dictare models rm all       # Remove all cached models

dictare config¶

Read and edit configuration.

dictare config list
dictare config get stt.model
dictare config set stt.language it
dictare config edit

dictare version¶

Show version and dependency information.

dictare version [options]

Option	Description
`--deps`	Show dependency versions

dictare version
dictare version --deps

dictare setup¶

Run the initial setup wizard. Guides you through audio device selection, STT model download, and hotkey configuration.

dictare setup

dictare tray¶

Manage the system tray icon.

dictare tray start      # Start tray icon
dictare tray stop       # Stop tray icon
dictare tray status     # Show tray status

The tray icon shows engine status, lets you toggle listening, switch agents, and access settings. It is installed automatically by the Dictare installer.

dictare dependencies¶

Check or resolve system dependencies.

dictare dependencies check
dictare dependencies resolve