About

Greeklish transliteration maps Greek Unicode code points to Latin-alphabet equivalents. The process is not a simple one-to-one substitution. Greek contains digraphs (ΜΠ → B, ΝΤ → D, ΓΚ → G) and diphthongs (ΑΙ → AI, ΟΥ → OU, ΕΙ → EI) that require greedy left-to-right matching before single-character fallback. Context matters: word-initial μπ yields b, while mid-word μπ yields mb. Incorrect transliteration corrupts search indexing, breaks URL slugs, and renders Greek names unrecognizable in Latin-script systems. This converter processes all 24 base letters, 7 accented vowels, polytonic variants, and the terminal sigma (ς → s) in a single pass.

The tool approximates the ISO 843 / ELOT 743 standard but follows the informal "internet Greeklish" convention where phonetic readability is prioritized over strict reversibility. Note: Greeklish is inherently lossy. Multiple Greek characters can map to the same Latin output (η, ι, υ all commonly map to i), making reverse conversion ambiguous.

Formulas

The transliteration algorithm performs a greedy left-to-right scan of the input string. At each position i, it checks for a two-character digraph match before falling back to a single-character lookup.

{

output += digraph_map[s_i + s_i+1], if s_is_i+1 ∈ digraph_mapoutput += char_map[s_i], if s_i ∈ char_mapoutput += s_i, otherwise (passthrough)

Where s is the input string, i is the current scan index, digraph_map contains all two-character Greek combinations (size ≈ 30 entries), and char_map contains all single Greek characters including accented variants (size ≈ 90 entries). Time complexity is O(n) where n is string length. Each lookup is O(1) via hash map.

For context-sensitive digraphs, word-boundary detection uses the rule: position i = 0 or s_i−1 is whitespace or punctuation → word-initial form. Otherwise → mid-word form.

Reference Data

Greek	Greeklish	Name	Notes
Α / α	A / a	Alpha	Base vowel
Β / β	V / v	Beta	Modern pronunciation is /v/
Γ / γ	G / g (Y before e/i)	Gamma	Context-sensitive
Δ / δ	D / d	Delta	Sometimes "dh" in older systems
Ε / ε	E / e	Epsilon	Short e
Ζ / ζ	Z / z	Zeta	-
Η / η	I / i	Eta	Modern Greek: /i/ sound
Θ / θ	TH / th	Theta	Digraph output
Ι / ι	I / i	Iota	Base vowel
Κ / κ	K / k	Kappa	-
Λ / λ	L / l	Lambda	-
Μ / μ	M / m	Mu	Part of μπ digraph
Ν / ν	N / n	Nu	Part of ντ digraph
Ξ / ξ	X / x	Xi	Pronounced /ks/
Ο / ο	O / o	Omicron	Short o
Π / π	P / p	Pi	-
Ρ / ρ	R / r	Rho	-
Σ / σ / ς	S / s / s	Sigma	ς is word-final form
Τ / τ	T / t	Tau	-
Υ / υ	Y / y	Upsilon	Sometimes "u" in diphthongs
Φ / φ	F / f	Phi	Sometimes "ph" in older systems
Χ / χ	CH / ch	Chi	Digraph output
Ψ / ψ	PS / ps	Psi	Digraph output
Ω / ω	O / o	Omega	Same as omicron in Greeklish
ΜΠ / μπ	B / b (or MB / mb)	Mu-Pi digraph	B word-initial, MB mid-word
ΝΤ / ντ	D / d (or ND / nd)	Nu-Tau digraph	D word-initial, ND mid-word
ΓΚ / γκ	G / g (or NG / ng)	Gamma-Kappa digraph	G word-initial, NG mid-word
ΓΓ / γγ	NG / ng	Double Gamma	Nasal /ng/ sound
ΟΥ / ου	OU / ou	Omicron-Upsilon	Diphthong /u/
ΑΙ / αι	AI / ai	Alpha-Iota	Diphthong /e/
ΕΙ / ει	EI / ei	Epsilon-Iota	Diphthong /i/
ΟΙ / οι	OI / oi	Omicron-Iota	Diphthong /i/
ΑΥ / αυ	AV / av (or AF / af)	Alpha-Upsilon	AV before voiced, AF before voiceless
ΕΥ / ευ	EV / ev (or EF / ef)	Epsilon-Upsilon	EV before voiced, EF before voiceless

Frequently Asked Questions

There is no single official Greeklish standard. ISO 843 and ELOT 743 define formal romanization, but internet Greeklish diverges significantly. For example, β maps to "v" in modern phonetic Greeklish (reflecting pronunciation /v/) but to "b" in classical romanization. Similarly, η maps to "i" phonetically but "e" or "h" in older systems. This tool follows the modern phonetic internet convention, which is the most widely understood by Greek speakers online.

No. Greeklish is a lossy mapping. Multiple Greek characters collapse to the same Latin output: η, ι, υ, ει, and οι all produce "i" or "ei"/"oi". Reversing requires a dictionary-based approach with statistical disambiguation, which has roughly 85 - 92% accuracy depending on corpus. This tool performs forward conversion only.

Word-initial μπ represents the /b/ sound (e.g., μπαλα → bala), so it maps to "b". Mid-word μπ represents /mb/ (e.g., καμπανα → kampana), mapping to "mb". The same logic applies to ντ (D/ND) and γκ (G/NG). The converter detects word boundaries by checking if the preceding character is whitespace, punctuation, or the start of the string.

Yes. All polytonic diacritical marks (oxia, varia, perispomeni, psili, dasia, ypogegrammeni, dialytika) are stripped during conversion. Characters like ἀ (alpha with psili), ἁ (alpha with dasia), ᾶ (alpha with perispomeni) all resolve to "a". The converter handles the full Unicode range U+0370 - U+03FF (Greek and Coptic) and U+1F00 - U+1FFF (Greek Extended).

In modern Greek phonology, αυ is pronounced /av/ before voiced consonants and vowels, but /af/ before voiceless consonants and at word end. Similarly for ευ (/ev/ vs /ef/). Full phonological analysis would require lookahead to the following consonant. This converter uses the simpler convention: αυ → "av" and ευ → "ev" uniformly, which is the most common internet Greeklish practice.

All non-Greek characters pass through unchanged. Latin letters, numbers, punctuation, emoji, and whitespace are preserved exactly. This allows mixed-script text (e.g., Greek text containing English brand names or URLs) to convert correctly without corruption.