Orukeet local transcription walkthrough

This edited walkthrough presents recorded CLI output, not a real-time screen recording or speed comparison. The model and runtime were installed before the demonstrated command. The full setup is at https://oruk.ai/guides/orukeet-local-transcription.

0:00 — Orukeet runs locally. This recording used Python 3.12, Orukeet 0.1.1, native Q8 and Metal on an Apple M5 Max.
0:04 — Hear the original 4.05-second quickstart recording: “I am genuinely excited that this is finally working.”
0:10 — Run python orukeet-local.py sample.wav --installation installation.json. The script reads the saved receipt and local audio.
0:17 — The returned text is “I am genuinely excited that this is finally working.” Its segment runs from 0.00 to 3.84 seconds. Language is null. Segment timing is not word alignment or diarization; this one recording does not establish accuracy.
0:25 — A separate recorded missing-file check exits with code 1 and prints “Transcription failed: Audio file not found: missing.wav”. Check the filename and run the command with an existing recording.
0:32 — The linked guide includes the pinned installation instructions, complete runner, full output and verification hashes.

Original audio: https://oruk.ai/samples/oruk-quickstart.wav
The runner reads the installed model, runtime and recording from local files.