🦜 parakeet-cli

Fast local speech-to-text. 25 languages. ~18x faster than Whisper on Apple Silicon.

• CoreML on Apple Silicon — ~155x real-time via FluidAudio
• ONNX on CPU — cross-platform fallback, 3x faster than Whisper
• Any audio format — ffmpeg handles OGG, MP3, WAV, FLAC, M4A
• Zero Python — Bun + TypeScript, native Swift binary for CoreML

Quick Start

bun install -g @drakulavich/parakeet-cli
parakeet install          # CoreML on macOS arm64, ONNX elsewhere
parakeet audio.ogg        # → transcript to stdout

Usage

parakeet install                 # auto-detect backend
parakeet install --coreml        # force CoreML (macOS arm64)
parakeet install --onnx          # force ONNX (~3GB)
parakeet audio.ogg               # transcribe (language auto-detected)
parakeet --version

Stdout: transcript. Stderr: errors. Pipe-friendly.

Requirements

• Bun >= 1.3
• ffmpeg in PATH (ONNX backend only)
• ~3GB disk (ONNX models)

Supported Languages

🇧🇬 Bulgarian, 🇭🇷 Croatian, 🇨🇿 Czech, 🇩🇰 Danish, 🇳🇱 Dutch, 🇬🇧 English, 🇪🇪 Estonian, 🇫🇮 Finnish, 🇫🇷 French, 🇩🇪 German, 🇬🇷 Greek, 🇭🇺 Hungarian, 🇮🇹 Italian, 🇱🇻 Latvian, 🇱🇹 Lithuanian, 🇲🇹 Maltese, 🇵🇱 Polish, 🇵🇹 Portuguese, 🇷🇴 Romanian, 🇷🇺 Russian, 🇸🇰 Slovak, 🇸🇮 Slovenian, 🇪🇸 Spanish, 🇸🇪 Swedish, 🇺🇦 Ukrainian

Benchmark

MacBook Pro M3 Pro — 10 Russian voice messages:

	faster-whisper (CPU)	Parakeet (CoreML)
Total	35.3s	1.9s
Speed		~18x faster

Full results with transcripts: BENCHMARK.md

How It Works

parakeet audio.ogg
  ├── CoreML installed? → parakeet-coreml subprocess → stdout
  └── ONNX installed?   → ffmpeg → mel → encoder → decoder → stdout

• CoreML: Swift binary wraps FluidAudio + CoreML model
• ONNX: NVIDIA Parakeet TDT 0.6B v3 via onnxruntime-node

OpenClaw Integration

Drop-in replacement for OpenClaw voice processing — no API keys, runs locally.

{
  "tools": {
    "media": {
      "audio": {
        "enabled": true,
        "models": [{"type": "cli", "command": "parakeet", "args": ["{{MediaPath}}"], "timeoutSeconds": 120}],
        "echoTranscript": false
      }
    }
  }
}

License

MIT