Felo

Kaithi Script Transliteration Online

Convert Latin romanized text into Kaithi Unicode characters, or paste Kaithi to read it back in Latin letters. The conversion uses an explicit letter and digraph map with dependent vowel signs applied after consonants, and characters outside the map are preserved rather than dropped. Try it online.

What is Kaithi Script Transliteration Online?

Kaithi Script Transliteration Online is a conversational tool that converts romanized Latin text into Kaithi script characters and converts Kaithi text back into Latin romanization. Kaithi is a historical North Indian script used in legal, administrative and literary documents, and its Unicode block renders as distinct characters rather than as Latin letters. This agent performs character-level transliteration: it substitutes letters and digraphs according to a fixed map, then attaches dependent vowel signs and marks to the preceding consonant. It does not translate meaning between languages, and it does not produce font files.

The important difference from a bare two-box web page is transparency. Every substitution the agent makes comes from a stated map, unsupported characters are kept in place instead of vanishing, and the agent tells you which characters it could not map. If you paste Kaithi and ask for Latin, you get a romanized reading with the same honesty about gaps.

Key Features

Latin to Kaithi character conversion. Independent vowels, consonants, and digraph consonants such as kh, gh, ch, jh, th, dh, ph and bh are mapped to their Kaithi codepoints.
Dependent vowel signs after consonants. Vowel endings such as -e, -i, -o, -u, -ā, -ī, -ū, -ai and -au become the corresponding Kaithi signs attached to the preceding consonant, and candrabindu, anusvara, visarga and virama marks are supported.
Reverse reading. Paste Kaithi text and receive a Latin romanization built from the same map, with a note on anything ambiguous.
Preservation reporting. Characters outside the map, such as digits, punctuation or Latin letters you did not intend to convert, are reported rather than silently removed.
Copyable output blocks. Results arrive as plain text you can copy directly into a document, database or design file.

How to Use

Type or paste your romanized text and say what you want, for example: convert this to Kaithi script. If you paste Kaithi instead, say convert to Latin.
Read the converted block. The agent shows the Kaithi output first, then a short note listing any characters it left unchanged and why.
Ask for a refinement if needed, such as keeping spaces exactly as typed, converting a longer passage, or explaining a specific sign.

Example Use Cases

Kaithi lipi to Hindi reading aid. Paste a short Kaithi inscription or document fragment and ask for a Latin romanization so you can sound it out before consulting a Hindi speaker.
Romanized name rendering in Kaithi script. Convert a personal or place name written in Latin letters into Kaithi characters for a label, poster or study sheet.
Classroom demonstration of Kaithi script transliteration. Convert a sample sentence and ask the agent to explain which letters became independent vowels and which became dependent signs.

Tips for Best Results

Write vowel length deliberately: ā, ī and ū are distinct from a, i and u, and the map treats them differently.
Use the digraph spellings the map expects, such as kh, gh, ch, jh, th, dh, ph and bh, rather than ad hoc combinations.
If a character comes back unchanged, that is intentional preservation, not a failure; ask the agent to list unmapped characters if the note is unclear.

FAQ

What input format does it accept?
Plain text. Latin romanized text for forward conversion, or Kaithi Unicode text for reverse conversion. Very long passages are handled in labeled segments so nothing is truncated.

How accurate is the conversion?
It is exact for characters and digraphs in the stated map, and it applies dependent vowel signs and marks by rule whenever a literal hyphen precedes the vowel or mark, as in "k-e" becoming 𑂍𑂵; a vowel without that hyphen stays an independent vowel. Very long passages are split into labeled segments of roughly 500 characters at word boundaries, and every segment is converted. It is not a linguistic translation and does not resolve ambiguous romanization, such as whether a bare a is inherent or explicit.

How does this compare with some free text tools?
Some free text tools give you two boxes and no explanation of the mapping or of unmapped characters. This agent states the map, preserves unsupported characters, reports them, and supports the reverse direction with the same transparency. It does not claim scholarly historical accuracy beyond the configured character set.