How accurate is speech recognition in Hungarian?

Hungarian is usable but clearly harder than the big Western European languages: about 14 to 17 wrong words per 100 on read speech. In our October 2026 benchmark on a Mac with an M2 Pro, Parakeet TDT 0.6B v3 had a word error rate of 14.09 percent ¹ and Whisper large-v3-turbo had 17.21 percent ², on 100 Hungarian recordings from the FLEURS test set. We have no Apple SpeechTranscriber result for Hungarian.

Parakeet beats Whisper by about three points in Hungarian

Parakeet has the lower error rate and is also the faster engine. It transcribed the recordings about 22 times faster than real time ³, against about 15 times for Whisper ⁴. Both are fast enough that you will not wait on either. The speed dataset covers every language.

Our measurement of the engines in Hungarian

Lower is better in this table.

The transcripts behind each row are on the on-device benchmark page.

Published results agree with ours within two points

NVIDIA reports 15.72 percent for Parakeet on FLEURS Hungarian ⁵, a little above our 14.09 percent. OpenAI's paper gives 17.0 percent for the older Whisper large-v2 ⁶, almost the same as our turbo result. The Hungarian error rate is a stable property of these models, not a quirk of one run.

All published rows are on the published accuracy page.

Hungarian ranks nineteenth of 25 languages

Taking the best engine in each language, Hungarian is nineteenth of the 25 in the benchmark. It is far behind Italian (2.52 percent with Parakeet ⁷) and Polish (7.70 percent with Whisper ⁸). It sits level with Estonian (14.17 percent ⁹), another language with many word endings, and ahead of Lithuanian (17.75 percent ¹⁰).

Formatting adds a couple of points

Counting punctuation and capital letters as a reader sees them, Parakeet's text differed from the written sentences by 16.56 errors per 100 tokens ¹¹, and Whisper's by 19.28 ¹². The gap between engines stays about the same, so the ranking holds.

What to expect when you dictate in Hungarian

In a quiet room, expect roughly 1 wrong word in 7 with Parakeet, so plan to proofread every dictated passage. Errors are most likely on names, numbers, rare words and inflected endings, since Hungarian packs a lot of grammar into suffixes and a single wrong suffix makes the whole word wrong. Parakeet is the first engine to try, and it runs entirely on the Mac. If your dictation app lets you choose, pick it for Hungarian over Whisper. Speak in full phrases rather than single words, and keep a correction list if your app offers one. We found no published fact that confirms Hungarian as a built-in Mac or Windows dictation language, so check your system's language list.

Parakeet is the better choice for Hungarian, and checking takes ten minutes

Parakeet wins on both accuracy and speed here, so there is no trade-off to weigh. The margin between the two engines is only a few points, though, and names and technical words will behave differently from the read sentences in the test. Dictate one page of your usual text through each engine you can run and count the corrections. If the counts are close, take the engine that your app supports best. The published accuracy page shows how other models compare on similar tests.

What these numbers do not show

The recordings are people reading prepared sentences in a quiet room, scored on one Mac with 100 recordings. Spontaneous speech, noise, regional accents and domain vocabulary are not in the test, and we did not measure Apple's model. Word error rate counts a wrong ending as a whole wrong word, which makes it strict for an agglutinative language. For the measure itself, see the guide on how accurate speech recognition is.

Our guide to European languages in speech to text helps you choose an engine, and the languages overview covers other languages. For comparison, read the same question for Polish, Croatian and Romanian.