Об инструменте
Simple case converters only offer “Sentence case” or “Title Case”, which ruins names, brands and acronyms. Truecasing needs context: “apple” the fruit versus Apple the company, “us” versus the US. This tool uses a multilingual language model to recognize proper nouns, organizations and acronyms, and applies each language’s rules — German nouns are capitalized, Turkish dotted İ is handled with locale-aware Unicode functions. Only letter case may change; if the model touches anything else, the passage is processed with deterministic rules instead.
Как и все инструменты LanguageHelper, ИИ-восстановление регистра работает полностью в вашем браузере. Текст никогда не загружается, не хранится и не используется для обучения, поэтому подходит для конфиденциальных документов, данных клиентов и личной переписки. Не нужны ни аккаунт, ни установка, а после загрузки инструмент работает и офлайн.
При первом запуске загружается компактная ИИ-модель — около 320 MB для быстрой модели по умолчанию или 480 MB для модели наилучшего качества. Она сохраняется в браузере, поэтому следующие запуски занимают секунды и работают офлайн. Пока модель загружается, показывается мгновенный результат на основе правил; выберите движок «Правила», чтобы ничего не скачивать.
Как это работает
- Select the text language so locale rules (like Turkish İ/ı) are applied.
- Paste text that is all lowercase or ALL CAPS.
- Click “Fix capitalization”.
- Copy the corrected text or download it as .txt or .md.
Сценарии использования
- Rewrite ALL-CAPS e-mails, tickets and forum posts so they don’t read as shouting.
- Restore capitals in text exported from systems that lower-case everything.
- Fix lower-case notes and transcripts before publishing.
- Normalize product names and acronyms in data-cleaning pipelines.
Примеры
До
I MET JOHN IN LONDON ON MONDAY. THE USA IS BIG.
После
I met John in London on Monday. The USA is big.
До
frau müller wohnt seit 2019 in münchen.
После
Frau Müller wohnt seit 2019 in München.
Советы
Choose the correct language before running: German capitalizes all nouns, Turkish needs dotted and dotless i handled correctly, and English always capitalizes “I”. If your text mixes lowercase and capitals, the tool keeps existing capitals where they are already correct. For data cleaning, compare the result with Text Diff and spot-check brand names, which the model may not know.
Частые вопросы
What about languages without capital letters?
Chinese, Japanese, Korean, Thai, Arabic and Hindi scripts have no letter case, so for them the tool fixes what does have case: Latin brand names and acronyms inside the text (iphone → iPhone, youtube → YouTube), ALL-CAPS Latin names, and full-width letters (GITHUB → GitHub).
Does it handle Turkish and other special cases?
Yes. Case mapping uses locale-aware Unicode functions, so Turkish “i” becomes “İ” and “ı” becomes “I”, and German “ß” is preserved correctly.
Can it capitalize German nouns?
The AI engine is instructed to capitalize all German nouns, which a simple rule cannot do. The rules engine capitalizes a list of frequent nouns, days and months as a fallback.
Is anything sent to a server?
No. Both engines run locally. The AI model is cached in your browser after the first download, so later runs work offline.
Why did some words stay lowercase?
Rare names and ambiguous words may not be recognized, especially by the rules engine. Review the result and adjust any remaining names by hand.