Script converter: change the script, keep the words

No. Transliteration changes only the script, not the language. “नमस्ते” becomes “namaste” or “નમસ્તે”, but a Hindi sentence converted to Gujarati script is still Hindi, written in Gujarati letters. To read Gujarati you need a translator.

IAST and ISO 15919 are scholarly standards that use diacritics (ā, ṭ, ś). They differ mainly in details such as e/ē, o/ō, and anusvara (ṃ vs ṁ). ITRANS and Harvard-Kyoto are plain-ASCII keyboard schemes: you can type them anywhere, and capital letters carry meaning (A for ā, T for ṭ).

Tamil script has one letter for each group such as ka/kha/ga/gha, so Hindi “क ख ग घ” all become “க”. Converting Tamil to other scripts reads every such letter as the unvoiced one (ka, ta, pa), because the spelling does not say otherwise.

Yes: conjuncts (क्ष, ज्ञ), vowel signs, anusvara, visarga, candrabindu, digits and the nukta letters (ड़, ढ़, फ़) are mapped. Letters that exist in only one script are mapped to the closest equivalent where there is one, and left unchanged where there is none.

It removes the silent “a” that Hindi speakers drop (कमल is “kamal”, not “kamala”), writes ऐ/औ as ai/au and shows nasal sounds as n or m. Hindi names and Sanskrit words are not always spoken with the same rule, so treat it as a readable draft. It works one way, from Devanagari to Roman.

No. The conversion runs entirely in your browser; nothing is sent to a server. The last text you typed is kept only in this tab’s session storage.