person

Levenshtein distance calculator

Levenshtein distance is the minimum number of single-character edits (insertions, deletions, or substitutions) needed to turn one string into another. Paste both into the input pane separated by a line of ---; the tool runs the standard dynamic-programming algorithm and reports the edit count, a similarity percentage, and both lengths. The compute runs in your browser; nothing uploads.

Input
Line 1:1 LF cloud_done Saved locally
Result Levenshtein Distance
0 lines 0 chars

Edit distance with a similarity score

Levenshtein distance counts the minimum number of single-character operations (insert, delete, or substitute) that transform one string into another. The classic example is kitten to sitting: substitute k for s, substitute e for i, insert g; three edits, distance 3. The algorithm fills a 2D table using the recurrence min(prev+1, curr+1, prev_diag + cost), where cost is 0 for matching characters and 1 for substitutions.

Output reports four values: the edit count itself, a similarity percentage computed as (1 - distance / max(len_A, len_B)) * 100, and both string lengths in code units. The percentage is rounded to one decimal place. Identical strings score 100% with distance 0; entirely disjoint strings score 0% at distance equal to the longer length.

The split between text A and text B is a line containing exactly ---. Comparison is exact: case sensitive, whitespace included. Memory is linear in the shorter side via two rolling arrays; time is O(n × m). For pure word-level overlap see text similarity (Jaccard); for the largest unchanged span see longest common substring.

How to use levenshtein distance calculator

  1. 1Paste string A into the input panel, then a line with ---, then string B.
  2. 2The four-line output appears as you type: distance, similarity %, length A, length B.
  3. 3Click Copy to copy the result block, or Download to save it.
  4. 4Use the similarity % for fuzzy-match thresholds (over 80% is "very close" for short strings).
  5. 5For an inline marked-up view rather than a single number, switch to character diff.

Keyboard shortcuts

Drive TextResult without touching the mouse.

Shortcut Action
Ctrl FOpen the find & replace panel inside the input Plus
Ctrl ZUndo the last input change
Ctrl Shift ZRedo
Ctrl Shift EnterToggle fullscreen focus on the editor Plus
EscClose find & replace, or exit fullscreen
Ctrl KOpen the command palette to jump to any tool Plus
Ctrl SSave current workflow draft Plus
Ctrl PRun a saved workflow Plus

How the distance is computed

Classic Levenshtein recurrence

The algorithm fills a 2D table where dp[i][j] is the distance between the first i characters of A and the first j of B. The recurrence is min(dp[i-1][j] + 1, dp[i][j-1] + 1, dp[i-1][j-1] + cost), with cost = 0 when characters match and 1 otherwise.

Linear memory rolling arrays

Two arrays of length m + 1 stand in for the full matrix, so memory is linear in the shorter side. Multi-kilobyte inputs still complete in well under a second on a modern browser.

Similarity % derived from distance

Reported as (1 - distance / max(len_A, len_B)) * 100, rounded to one decimal place. Identical strings score 100%; an empty string compared to a non-empty string scores 0%.

Case and whitespace count

Comparison is exact. Hello versus hello registers a substitution. A space versus a tab also costs one. Lowercase or strip first if you want softer matching.

Three-hyphen separator

The split between A and B is a line containing exactly ---. The tool needs this marker to know which side is which.

Worked example

Three edits transform kitten into sitting: k -> s, e -> i, insert g. The 57.1% comes from (1 - 3/7) * 100. For an inline marker view see character diff.

Input
kitten
---
sitting
Output
Levenshtein distance: 3
Similarity: 57.1%
Length A: 6
Length B: 7

Settings reference

Behaviour Effect on output
Separator A line containing exactly --- splits text A from text B.
Algorithm Standard Levenshtein with rolling-array dynamic programming.
Operations counted Insert, delete, substitute (each cost 1).
Match rule Exact codepoint equality, case sensitive.
Similarity formula (1 - distance / max(len_A, len_B)) * 100, one decimal place.
Length unit UTF-16 code units (JavaScript .length).
Empty side Distance equals the length of the other side.

FAQ

What counts as one edit?
Inserting one character, deleting one character, or substituting one character. Each costs 1. Transposition (swapping two adjacent characters) counts as 2 in classic Levenshtein; for the transposition-aware variant use Damerau-Levenshtein.
How is similarity calculated?
As (1 - distance / max(len_A, len_B)) * 100, rounded to one decimal place. So distance 3 against a 7-character string gives (1 - 3/7) * 100 = 57.1%.
Is the comparison case sensitive?
Yes. Hello versus hello costs one substitution. Lowercase both sides first with lowercase if you want case-folded distance.
How does this differ from Jaccard similarity?
Jaccard similarity works on word sets and ignores order. Levenshtein works at character level and respects order. Levenshtein for short strings (names, codes), Jaccard for paragraphs.
How long are the inputs allowed to be?
There is no hard cap, but time is O(n × m), so a 10,000 by 10,000 character compare takes noticeable time. For very long inputs prefer longest common substring or word-level similarity.

Also known as

levenshtein distance edit distance calculator string distance tool fuzzy match score text similarity edits compare strings distance minimum edit count spell distance