LLM on Mac / Model

Whisper large-v3-turbo

MLX / FP16

Starting memory estimate

16 GB

For trying short tasks one at a time

The official ~6GB figure describes GPU memory, not total Mac RAM. The 16GB starting estimate includes the OS and audio app. Split recordings and run the summarizer separately.

Tasks to try

Transcribe recordings. For meeting notes, pass the transcript to a text model such as Qwen3.5-9B afterward.

Keep under review

Speaker diarization is a separate feature. Check noise, overlapping speech and text invented during silence.

Getting started

Memory figures are editorial budgets including room for the OS and apps, not verified minimums. Text models assume 4-bit weights, 4K–8K input tokens and one model at a time. Long documents, images and concurrent apps need more. A3B does not mean the entire model fits in 3GB.

Assumes an Apple Silicon Mac and a compatible app: LM Studio or similar for text, Whisper apps for audio, and model-specific image/video tools. RAM alone does not determine speed. Start with a short input on your current Mac.

Measurements for this model

2026-11-06 JST · Mac mini · mlx-whisper 0.4.3

Processing time18.17 s
Character error rate (CER)10.5%
Peak memoryNot measured

Result: failed

A transcript was produced, but the exact-match checks for 十時 and 三部 failed. This is one task including numeral-format differences, not an overall accuracy rating. Peak memory was not sampled.

Test conditions
{
  "language": "ja",
  "task": "transcribe",
  "temperature": 0,
  "word_timestamps": false,
  "model": "mlx-community/whisper-large-v3-turbo",
  "revision": "a4aaeec0636e6fef84abdcbe3544cb2bf7e9f6fb",
  "platform": "macOS-27.2-arm64-arm-64bit",
  "input": "voice-asr-synthetic-ja-v1"
}

Actual transcript

明日の会議は10時からです。資料は3部用意してください。メールはベンチマークのexample.comに送ってください。