On standard American English, the difference is real but not dramatic. As a family, larger Whisper models generally post lower English WER than smaller ones, so large-v3 tends to sit below medium on clean English — but OpenAI does not publish a single tidy per-model, per-language WER table I can quote a specific number from, and I'm not going to invent one. Treat the direction (large-v3 < medium on English) as the reliable part, and the exact margin as something you should measure on your own audio.

That gap is meaningful for production dictation. At a low error rate you mostly see punctuation slips and the occasional homophone ("their" vs "there"). As the error rate climbs you start losing content words, especially names and technical terms. Medium is not broken — on accented or noisy input it's just noticeably noisier than large-v3.

The catch: "standard American English in a quiet room" is a narrow slice of what people actually dictate. Move the speaker, the mic, or the accent, and the gap widens.

Honest answer: nothing per-accent. OpenAI evaluates Whisper on whole-language benchmarks — FLEURS (102 languages), VoxPopuli (14 European languages), Common Voice, and LibriSpeech. None of these are broken down by "Indian English," "Brazilian Portuguese-accented English," or "Cantonese-accented English." That's a real gap in the public literature, and I'm not going to paper over it with numbers I made up.

What we can infer is qualitative. Accents tend to degrade WER because they introduce phonetic patterns the model saw less of during training. Models with more parameters and more training data — i.e., large-v3 — generally handle those patterns better. That's the pattern across the Whisper family, but it's a general expectation, not a per-accent number from OpenAI.

So the honest picture is: large-v3 probably beats medium more on accented English than on standard English, by a margin we can't quantify from OpenAI's published charts alone. That's the truth. If a vendor claims a specific percentage improvement on "Indian accent," ask for the test set.

This is where model size tends to matter most, and it matters more than the accent question for many users. The general pattern for larger Whisper models is that they hold up better on lower-resource languages — the gap over a smaller model tends to widen as you move from high-resource languages (English, Spanish, French) toward lower-resource ones (Swahili, Welsh, Lao), where there was simply less training data to go around.

On a low-resource language, a smaller model's errors can compound to the point where the transcript needs heavy cleanup, while large-v3 stays usable. If you dictate in anything other than English — even occasionally — that tendency is the strongest argument for large-v3.

This is also why I default to the largest model that fits my hardware in my offline dictation setup. MetaWhisp itself ships with large-v3-turbo specifically because it preserves most of large-v3's accuracy on non-English languages while running in roughly half the time on the Neural Engine.

Noise is where the model-size gap tends to show up most clearly in practice. Larger Whisper models generally degrade more gracefully as recording conditions get worse, and large-v3 also tends to hallucinate less on silence and background hum — the famous Whisper problem where the model invents a sentence during a quiet stretch.

Medium hallucinates more. If you've ever finished dictating and found a phantom "Thanks for watching!" at the end of your transcript, that's medium's classic failure mode. Large-v3 doesn't eliminate it, but it's rarer and usually shorter when it does happen.

For long-form dictation — meetings, lectures, podcasts — hallucination matters more than a 2% WER improvement. Every hallucinated sentence is cleanup work. This alone pushes the value proposition toward large-v3 for anything beyond quick voice notes.

Yes, with caveats. Whisper large-v3 runs on Apple Silicon via WhisperKit, and on an M1 or later it will produce a transcript — but the latency depends heavily on the chip and audio length. On an M1 Air with the official Core ML port, large-v3 runs slower than medium — a 60-second clip takes a couple of minutes to transcribe.

An M2 Pro or M3 Pro cuts that significantly. On M4 Pro and Max chips, large-v3 feels nearly real-time for short clips. For an M1 Air user who dictates a lot, large-v3-turbo is the practical sweet spot — roughly half the latency of large-v3 with only a small accuracy trade-off on most languages.

If your workflow is "hold hotkey, dictate three sentences, paste into email," any current Apple Silicon Mac handles large-v3 fine. If you're transcribing an hour-long meeting, you'll want either a Pro chip, large-v3-turbo, or a cloud pass to skip the wait.

large-v3-turbo is OpenAI's distilled variant of large-v3 — fewer decoder layers, similar encoder, same general accuracy tier. It launched in late 2024 and is what most production apps now ship by default. The tradeoff it makes: a small accuracy drop for a large speedup, often 2x or better on the same hardware.

For accented English specifically, large-v3-turbo is the option I usually recommend. It keeps nearly all of large-v3's gains over medium while running in the time budget of an everyday dictation app. MetaWhisp uses it for exactly this reason — local mode is free and unlimited on macOS 14+ Apple Silicon, and the accuracy hit versus the full large-v3 is small enough that most users don't notice.

I wrote a deeper dive on large-v3-turbo vs full large-v3 if you want the model-level details.