Find and replace tool

This tool locates every occurrence of a specified string within a block of text and replaces each one, in a single pass. The matching rule is literal and case sensitive, with no whole-word boundary or pattern matching applied, and it's worth being exact about what that means before relying on it.

Processing runs entirely client-side, in JavaScript, within your browser. There is no upload and no server round trip, and accordingly no run limit.

Two fields to set.

Step 1

Paste the text and enter the string to find.

Step 2

Enter the replacement string.

Step 3

Run the replace and read the result.

This tool matches the exact character sequence entered, wherever it occurs, including inside a longer word. A search for 'cat' matches the 'cat' within 'concatenate' identically to a standalone occurrence of 'cat', since no word-boundary check constrains the match.

Where a whole-word match is required, that constraint must be built into the search string itself, a leading or trailing space, or the punctuation expected around the target, to prevent a match landing inside an unrelated longer word.

This tool matches capitalization exactly. A search for 'Report' will not match 'report' or 'REPORT'. Where text carries inconsistent capitalization of the same term and every variant needs to be caught, run the tool once per capitalization, or normalize case first with the case converter and search against a single consistent form afterward.

This is a deliberate design choice rather than a limitation: case-sensitive matching is predictable and never alters a capitalization outside the search scope, at the cost of requiring multiple passes against inconsistently capitalized text.

Typing what is visible on screen only produces a match if the typed character and the character present in the text are the same codepoint. A straight apostrophe typed on a keyboard is a distinct character from a curly apostrophe introduced by a copy and paste, and a search for one will never match the other, despite near-identical rendering.

The same applies to hyphens versus en dashes versus minus signs, and to a regular space versus a non-breaking space. When a find-and-replace fails to match text that is clearly visible, the typed character and the actual character are most likely different codepoints rendering identically. Copying the exact target text directly from the source, rather than retyping it, avoids this failure mode entirely.

A rename catching only some occurrences leaves a document referencing two different names for the same entity. A search failing silently due to a look-alike character can lead to the false conclusion that a phrase does not appear in a document, when it is present repeatedly under a slightly different encoding.

In bulk text edits, an overly broad match lacking whole-word constraints can silently corrupt words that happen to contain the search string as a substring, an error easily missed on a cursory review.

Editorial teams swap a name, term or phrase consistently across a long document. Engineers update repeated boilerplate text, a company name, a version string, a placeholder, in content outside a codebase with its own find-and-replace tooling.

It is also used to remove a recurring stock phrase from AI-generated drafts, or to correct a term consistently misspelled or mis-capitalized throughout a document.