March 28, 20263 min read

Best Whisper Model for Dictation — Comparing Speed vs. Accuracy

Whisperer ships with multiple Whisper models. This raises an obvious question: which one should you actually use?

Short answer: probably Turbo. Longer answer: it depends on what you're doing.

The models#

ModelSizeRAM UsageSpeed*AccuracyBest For
Tiny~75MB~200MBInstantFairQuick notes, low-end hardware
Base~150MB~350MBVery fastGoodFast dictation, older Macs
Small~500MB~800MBFastVery goodBalanced everyday use
Medium~1.5GB~2.5GBModerateExcellentHigh-accuracy dictation
Large V3~3GB~5GBSlowerBestMaximum accuracy, long dictation
Turbo~800MB~1.5GBFastVery goodBest speed/accuracy balance

*Speed tested on M1 MacBook Pro, 30-second recording.

My recommendations#

Most people: Turbo

Tip

Turbo is fast and accurate enough for most dictation. It's noticeably quicker than Large V3 without giving up much accuracy for English. Start here.

Maximum accuracy: Large V3

If you have 16GB+ RAM and need the best possible accuracy, Large V3 wins. Especially useful for:

  • Non-English languages
  • Heavy accents
  • Technical terms
  • Noisy rooms

Speed over accuracy: Small or Base

Small gives you near-instant results with occasional mistakes. Base is even faster but noticeably less accurate with complex speech.

Older Macs: Base or Tiny

Intel Macs or 8GB Apple Silicon should stick to Base or Small. Bigger models will choke your system.

Language matters#

Info

Whisper handles English well even at smaller sizes. Other languages need more horsepower:

  • Spanish, French, German — Small or Turbo work fine
  • Chinese, Japanese, Korean — use Medium or Large V3
  • Everything else — Large V3 is your safest bet

Preloading#

Enable model preloading in Settings to skip the 1-3 second cold start. The model stays in RAM, ready to go. Only do this if you have RAM to spare.

Switching models#

You can swap models without restarting. I use Turbo for quick notes and switch to Large V3 when accuracy matters more than speed.

Tips#

Close other apps

Large V3 on 16GB? Close Chrome and its 47 tabs first.

Turbo for live preview

Turbo is fast enough for real-time feedback while you speak.

Preload your model

Keeps the model in RAM so there's no startup delay.

Pick your language

Selecting a language explicitly beats auto-detect for accuracy.

Get a decent mic

Input quality matters more than model size. A $30 USB mic beats a laptop mic with Large V3.

Short version#

Start with Turbo. Not accurate enough? Try Large V3. Need it faster? Drop to Small. You can download multiple models and switch whenever.

Related: Whisper vs NVIDIA vs Apple Speech, Offline Transcription Engines, Getting Started Guide. See pricing and all features.

Ready to try voice dictation on your Mac?

Free download. No account required. 100% offline.

Download on the Mac App Store

Related articles

Try it.

Pay once. Keep it forever. Nothing goes to the cloud.

Free trial included. Pro Pack $14.99 lifetime.