Does it reformat my citations into a style?

No. It preserves your existing wording and fixes structure: splitting entries, removing duplicates, normalising whitespace, and sorting. To convert between styles, format from source data in the citation generator or bibliography builder instead.

How does it tell entries apart?

It splits on blank lines, on list markers like 1. or [3], and on the hanging-indent pattern where a flush-left line follows an indented one. That covers how reference lists are usually pasted from documents and PDFs.

How are duplicates matched?

By a normalised comparison that lowercases the text, unifies smart quotes and dashes, strips trailing punctuation, and removes other punctuation. So entries that differ only in those cosmetic ways are treated as the same and collapsed to one.

Is the style detection reliable?

It is a heuristic based on tell-tale shapes, such as a parenthesised year early in the entry for APA, so it is a helpful hint rather than a guarantee. If it misreads your list, set the style manually with the override.

What does normalising whitespace change?

It collapses runs of spaces and tabs into single spaces and removes spaces that sit before punctuation like full stops and commas. It does not touch the words themselves, so your citations read the same, just without the stray gaps.

FREE | SORT | DEDUPE | APA MLA CHICAGO HARVARD

Bibliography cleanup

Paste a messy bibliography. Get a cleaned, deduplicated, alphabetically-sorted version. Works on APA, MLA, Chicago, and Harvard mixes.

Detected style

Sort

Dedupe

Normalize whitespace

Original entries

Cleaned entries

Duplicates removed

Original

Cleaned

Cleaned bibliography appears here.

How the cleanup works

Entries are split on blank lines, hanging-indent boundaries, and numbered list markers. Each entry is normalized for whitespace and quote characters, then compared against the others to find duplicates. The detected style is a heuristic guess based on patterns like (Year) for APA and Harvard, quoted titles for MLA, and the Author. Year. sequence for Chicago.

Nothing is uploaded. Everything happens in your browser. Copy the cleaned output or download it as a .txt file.

About the Bibliography cleanup

The bibliography cleanup tool tidies a messy reference list you already have. Paste raw text, however it is separated, by blank lines, hanging indents, or numbers, and it splits the list into individual entries, removes duplicates, normalises spacing, and sorts the result. It detects the dominant style, APA, MLA, Chicago, or Harvard, and reports counts so you can see what changed.

It is for the common situation where references arrive in a jumble: copied from several documents, pasted with inconsistent whitespace, or repeated by accident. Rather than reformatting each citation, it preserves your wording and fixes the structural mess around it. Everything runs in your browser as you adjust the options, with the cleaned list shown side by side with the original.

How to use it

Paste your messy reference list into the Original box. Entries can be separated by blank lines, hanging indents, or list numbers.
Let the tool auto-detect the style, or override it with the Detected style selector if it guesses wrong.
Choose a sort order: alphabetical, by year, or grouped by source type.
Keep Dedupe on to drop repeated entries, and Normalize whitespace on to collapse stray spaces and fix spacing before punctuation.
Read the stats, original entries, cleaned entries, and duplicates removed, to confirm the result matches what you expected.
Copy the cleaned list or download it as a .txt file.

Examples

Splitting a list with no blank lines

Paste a reference list where each entry uses a hanging indent but no blank line between them. The tool recognises that a line starting at the left margin after an indented line begins a new entry, so it separates them correctly instead of merging the list into one block.

Removing an accidental duplicate

If the same source was pasted twice with slightly different trailing punctuation, the dedupe step compares a normalised version of each entry, ignoring case, smart quotes, and final punctuation, and keeps only the first. The duplicates-removed stat ticks up to one so the change is visible.

Grouping by source type

Set sort to By source type and the tool orders journals first, then books, then web sources, using cues like volume(issue) and pp. for journals and Press or Publishing for books. Within each group entries stay alphabetical, giving a list organised by kind of source.

Frequently asked questions

Does it reformat my citations into a style?: No. It preserves your existing wording and fixes structure: splitting entries, removing duplicates, normalising whitespace, and sorting. To convert between styles, format from source data in the citation generator or bibliography builder instead.
How does it tell entries apart?: It splits on blank lines, on list markers like 1. or [3], and on the hanging-indent pattern where a flush-left line follows an indented one. That covers how reference lists are usually pasted from documents and PDFs.
How are duplicates matched?: By a normalised comparison that lowercases the text, unifies smart quotes and dashes, strips trailing punctuation, and removes other punctuation. So entries that differ only in those cosmetic ways are treated as the same and collapsed to one.
Is the style detection reliable?: It is a heuristic based on tell-tale shapes, such as a parenthesised year early in the entry for APA, so it is a helpful hint rather than a guarantee. If it misreads your list, set the style manually with the override.
What does normalising whitespace change?: It collapses runs of spaces and tabs into single spaces and removes spaces that sit before punctuation like full stops and commas. It does not touch the words themselves, so your citations read the same, just without the stray gaps.

Good to know

This tool deliberately leaves your citation wording alone. That is the right choice when the references are already in a style and only the surrounding mess needs fixing, but it means it cannot rescue a citation that is formatted wrong in the first place. If an entry has the author or date in the wrong place, cleanup will keep that mistake; rebuild it from source data in a formatting tool instead.

The style and source-type detection rely on patterns, not a full parser, so unusual or heavily abbreviated references can be misread. Use the manual style override when the auto-guess is off, and skim the cleaned output rather than trusting the counts alone, since a near-duplicate that differs in more than punctuation will be kept as a separate entry. For lists you are building yourself rather than inheriting, the bibliography builder gives more control over format and order from the start.