Italian translation and speech-to-text models it

29 models into Italian, 16 from Italian, 14 speech-to-text. Best scored into Italian: English to Italian (chrF++ 56.4). Best scored from Italian: Italian to English (chrF++ 57.9).

Which open models work for Italian?

Windy Translate lists 29 open translation models into Italian and 16 from Italian. All of them can be downloaded from Hugging Face and run on your own hardware. By licence: 41 Apache-2.0 and 4 CC-BY-4.0 (terms and attribution text: licensing). 28 of them have a screening score on FLORES-200. The best scored into Italian is translate-tc-big-en-it (chrF++ 56.4, Good). The best scored from Italian is translate-it-en (chrF++ 57.9, Good).

With English. English to Italian: translate-tc-big-en-it (chrF++ 56.4, Good), the best of 2 direct models. Italian to English: translate-it-en (chrF++ 57.9, Good), the best of 2 direct models. To translate between Italian and a language with no direct model, you can go through English with these two.

By band, the scored Italian models are 14 Good, 12 Usable, 1 Limited and 1 Not recommended (Excellent is chrF++ 60 or more, Good 45 to 60, Usable 30 to 45, Limited 15 to 30). A score is a 48-sentence screening of news and encyclopedia text, not a measure of accuracy, and a difference of a point or two is noise. How we measured.

Speech-to-text. 2 speech-to-text models made for Italian: listen-windy-lingua-it and listen-windy-lingua-it-ct2. Also, Whisper's tokenizer lists Italian as one of its languages, so all 12 multilingual Whisper-based models here can transcribe it: see the speech models. Accuracy differs a lot between languages and model sizes, and this site has not measured speech accuracy.

Each model page has a short snippet to download and run the model yourself, its licence and attribution text, and its score where there is one. Pick the best scored model for your direction, then try it on a few sentences of your own text: a screening score is a starting point, not a guarantee.

Into Italian

From Italian

Italian speech-to-text

Scores: chrF++ on FLORES-200 dev, 48 sentences, measured 2026-08-04. Method.