Transliterate text between Indic scripts and Roman schemes such as IAST and ITRANS. It runs on your device, so your text is never sent anywhere.
Script Conversion rewrites text from one writing system into another. It does not translate. The sounds are carried across and the meaning is left alone, so Hindi written in Roman letters is still Hindi. A reader who does not know the language can now pronounce it. A reader who does know it can read it in a script they are used to.
The conversion runs on your device, using the open source Sanscript library under the MIT license. Nothing you type is sent to a server.
Eighteen schemes are offered. Eleven are scripts: Devanagari, Bengali, Gujarati, Gurmukhi, Kannada, Malayalam, Odia, Tamil, Telugu, Grantha and Sinhala. Seven write the same sounds in Roman letters. IAST and ISO 15919 use marks above and below the letters. ITRANS, Harvard-Kyoto, SLP1, Velthuis and WX use plain keyboard characters only.
Characters the library does not recognize pass through unchanged, so punctuation, digits and words in other alphabets survive a conversion.
Set the input scheme to ITRANS and the output to Devanagari. Type namaste and the output is नमस्ते. Press the swap button and the output becomes the input. Devanagari to IAST returns namaste, so this word survives the round trip unchanged.
The Devanagari संस्कृतम् is written saṃskṛtam in IAST, saṁskr̥tam in ISO 15919, saMskRtam in Harvard-Kyoto, saMskftam in SLP1, saMskRRitam in ITRANS, and সংস্কৃতম্ in Bengali.
The two schemes with marks disagree on one vowel. IAST writes it ṛ and ISO 15919 writes it r̥. Both describe the same sound.
No. It changes the letters, not the words. namaste becomes नमस्ते, which is the same word in a different script. To find out what an English word means, use the Dictionary.
The schemes are writing systems rather than languages, and one script serves many languages. Devanagari is used for Hindi, Marathi, Nepali and Sanskrit. Bengali script is used for Bengali and Assamese. Gurmukhi is used for Punjabi.
Because some scripts write fewer distinctions than others. Tamil has one letter, க, where Devanagari has four: क, ख, ग and घ. All four give க. Converting க back has to choose one, so all four return as घ. The word भारत goes to Tamil as பாரத and comes back as भारध.
No. The library runs in your browser. There is no account, nothing is uploaded, and the text is gone when you close the tab.
Yes. The application is stored in your browser after your first visit, and the conversion needs nothing else.
Choose IAST or ISO 15919 to be read by a person, because printed books use those forms. Choose ITRANS, Harvard-Kyoto, SLP1, Velthuis or WX to type on a plain keyboard, because each uses ordinary characters only.
Transliteration is reversible only when both scripts record the same distinctions. Devanagari to IAST is a one-to-one map, so text can go out and come back unchanged. Devanagari to Tamil is not, because Tamil script does not mark the difference between an unvoiced, aspirated, voiced and voiced aspirated consonant. That difference is not lost in Tamil, where a reader supplies it from the word being read. It is lost in the conversion, because no letter is left to carry it.
SLP1 and WX were designed for machines rather than readers. Each uses exactly one ASCII character for each sound, so a program can count, sort and compare letters without decoding a multi-byte script. This is why saMskRRitam in ITRANS becomes saMskftam in SLP1, where a single f stands for the vowel IAST writes as ṛ.