---

name: transcribe description: Speech-to-text with local NVIDIA Parakeet (NeMo) via the transcribe CLI. metadata: {"clawdbot":{"emoji":"🎙️","requires":{"bins":["transcribe"]}}}

Transcribe

Local speech-to-text on NixOS (Parakeet / NeMo). GPU auto when CUDA available.

Quick start

Copytranscribe /path/to/audio.ogg
transcribe /path/to/audio.wav --output /tmp/out.txt

Notes

• Default model: nvidia/parakeet-tdt-0.6b-v3
• Device auto: CUDA if available, else CPU
• Formats: any ffmpeg-supported input (ogg/wav/mp3/m4a)
• Former name: parakeet-transcribe

Useful env / flags

CopyPARAKEET_MODEL=nvidia/parakeet-tdt-0.6b-v3 transcribe file.ogg
PARAKEET_QUIET=0 transcribe file.ogg
PARAKEET_VENV=~/.local/share/parakeet-venv-py312 transcribe file.ogg
transcribe --device cuda file.ogg