Alternatives

Whisper / whisper.cpp

• Pros: high accuracy, many languages, runs offline with the right model.
• Cons: requires downloading a model, larger memory footprint, not the native macOS ASR path.
• Use when: you need state-of-the-art accuracy or a language Apple does not support well.

macOS Dictation

• Pros: built-in, keyboard-triggered, uses the same Speech framework.
• Cons: GUI-focused, not scriptable, no JSON output, no continuous keyword mode.
• Use when: you just want to dictate text into an app.

Siri

• Pros: deeply integrated, can perform system actions.
• Cons: no CLI or third-party automation API for arbitrary commands.
• Use when: you want Apple’s built-in assistant experience.

Web APIs (OpenAI, Google, etc.)

• Pros: often more accurate, supports many languages.
• Cons: requires network, API keys, uploads audio to a third party, latency and cost.
• Use when: accuracy or language coverage is more important than privacy/latency.

When to choose maclisten

Choose maclisten when you want:

• A tiny native binary with no model download and minimal resource use.
• Audio processed locally, with an --on-device option that keeps it on your Mac.
• Compact JSON output for scripts and agents.
• Continuous keyword listening on macOS.
• Audio and transcript saved together.