Does APA 7 still use “Retrieved from”?

No. Give the bare URL with no lead-in and no final full stop. A retrieval date appears only for pages designed to change, such as a wiki or a live feed.

What if the web page has no author?

The title moves into the author position and the date follows it. Check first whether an organisation is the author — on institutional pages it almost always is.

Yes, and it stays free: no credits, no account, no ads, no watermark, and the BibTeX and RIS downloads cost nothing either. Fifty sources a day is the only limit, and it exists to stop scripts, not students.

Which styles are supported?

APA 7th, MLA 9th, Chicago 17th (both the note and the author-date forms) and Harvard as set out in Cite Them Right. All four are generated from one lookup.

Can I paste a DOI or an ISBN instead of a link?

Yes, and you should when you have one. A DOI goes to Crossref and comes back with the journal, volume, issue and pages already correct — far better metadata than any article page publishes.

Why is the author missing or wrong?

Because that is what the page published. Many sites list no author, or list a staff account. The output says when no author was found; replace a staff byline with the person named on the page, or with the organisation.

Why does my reference say n.d.?

No publication date was found in the page's metadata. Check the footer, the byline and any “last updated” line, and add it by hand — n.d. is correct only if the source genuinely carries no date.

Does it fix the capitalisation of titles for APA?

No, on purpose. Re-casing a title automatically also lower-cases proper nouns, which introduces errors that look like yours. The title comes through as published and you adjust it where you can see the names.

Can it cite a book chapter or an edited collection?

Yes — set “Source is a” to book chapter. Detection cannot tell a chapter from a book from metadata alone, and the styles treat them very differently.

Can I look a book up by its title?

Yes, if you do not have the ISBN: type the title and the closest match in Open Library is used. It is a guess, so the result names what it matched and asks you to check it — the ISBN from the back cover is always more reliable.

What about a paywalled or login-only page?

Nothing here can see behind a paywall, so the lookup fails and says so. Cite it from what you can see, or find the article's DOI, which works regardless of access.

Is this accurate enough to submit?

The formatting is; the metadata is only as good as what the source published. Read the reference once against the page — ten seconds — and fix the author or date the warnings flag.

Do you store what I cite?

The citation stays attached to your job so you can come back to it, and nothing is published or shared. No account, no email.

APA Website Citation Generator — APA 7 format for a web page

The APA 7 rules for a web page

The shape is: Author. (Year, Month Day). Title of the page in italics. Site Name. URL

Four things changed in the 7th edition and still trip people up.

"Retrieved from" is gone. APA 7 wants the bare URL, with no lead-in phrase and no full stop after it. A retrieval date is now only used for a page designed to change under you — a wiki, a live dashboard, a social media profile — and then it reads "Retrieved March 14, 2026, from URL".

The site name goes after the title, not before it, and it is not italicised. The page title is.

When an organisation wrote the page, the organisation is the author. If the author and the site name would be identical — a page on the WHO site written by the WHO — you give the author and omit the site name entirely.

And a page with no date really does read (n.d.). Markers see it as a gap in your sourcing rather than a formatting choice, so check the footer and any "last updated" line before you accept it.

What this does

Paste the address of a page, a DOI, or a book's ISBN. The source is looked up, and you get the reference-list entry and the in-text citation in APA 7, MLA 9, Chicago 17 and Harvard — all four, every time, because the person who needs APA today needs MLA for the same source next term.

Underneath each reference is what most generators hide: a list of what was found and what was not. A missing date or author changes the reference in every style, and you cannot fix what you were not told about.

DOIs go to Crossref, which is where publishers register their metadata, so journal articles come back correct including volume, issue and page range. ISBNs go to Open Library, and so does a book title when the ISBN is not to hand — a title is a guess, and the result says so. Web pages are read from their own metadata.

Why references come out wrong, and how to spot it

Every citation generator is only as good as the metadata it was given, and web pages publish metadata carelessly.

The four failures worth knowing:

No author. Blogs and institutional pages often publish none, so the entry starts with the title — correct only if the page really is anonymous. Check for a byline, and use the organisation if there is one.
No date. You get n.d., which markers notice. It is usually in the footer, under the headline, or in a "last updated" line.
The wrong title. Sites put their own name in the page title. It is stripped where it can be recognised; check the title matches the heading on the page.
A staff byline. "Guardian Staff" or "Admin" is metadata, not an author.

None of this is exotic. It is the normal condition of the web, which is why the output tells you what it did not find.

Capitalisation is left exactly as published

APA wants sentence case for article and book titles. MLA and Chicago want title case. This tool changes neither, and that is deliberate.

A machine that re-cases titles also lower-cases proper nouns: "The Rise of Fascism in Europe" becomes "The rise of fascism in europe", and a marker sees a mistake you did not make. Re-casing correctly requires knowing which words are names, which is exactly what a formatter does not know.

So the rule here is: the title arrives as the source published it, and if your style wants a different case, you change it in your document — where you can see the proper nouns. For APA that usually means lower-casing everything after the first word and after a colon, leaving names capitalised.

In-text citations, which is where marks are actually lost

A perfect reference list with wrong in-text citations still loses marks, so each style shows both forms: the parenthetical one, and the narrative one you use when the author is the subject of your sentence.

The differences are small and consistent. APA puts a comma before the year and an ampersand between two authors: (Smith & Jones, 2020). MLA uses no comma and a page: (Smith 45). Chicago author-date drops the comma before the year but keeps one before the page: (Smith 2020, 45). Harvard follows APA's commas but writes "and": (Smith and Jones, 2020, p. 45).

Three or more authors collapse to "et al." in every style, from the first citation.

Page numbers belong in the in-text citation, not the reference, for anything you quote directly: the reference gives the whole work, the citation says where you were looking.

What to do about Harvard

Harvard is not a style. There is no Harvard manual, no committee and no edition number — it is a family of author-date conventions, and universities each publish their own variant.

What this produces follows Cite Them Right, which is what most UK institutions mean by Harvard and what most university libraries hand out: the year in brackets after the author, single quotes round article titles, italics on book and journal titles, and "Available at: … (Accessed: …)" at the end of a web reference.

If your department publishes its own guide, that guide wins over anything here — and the differences are usually punctuation rather than substance. Copy the reference, then match its commas to your handout once; every later reference follows the same pattern.

Exports, and using a reference manager instead

Every source downloads as BibTeX or RIS as well as text. BibTeX is what LaTeX and Overleaf read; RIS is what EndNote, Mendeley and library catalogues speak. Both are free, because a free tool that will not let you take your data out is not free.

For anything longer than an essay, use a reference manager. Zotero is free and open source, stores a source once and re-formats it into any style forever: it is the right tool for fifty sources, where a generator is the right tool for five.

This page is built for the five — the essay due tomorrow, the article a supervisor sent by link — without an account, without ads, and without emailing you.

Example result

YOUR SOURCE

10.1038/nature12373

YOUR RESULT

APA: Kucsko, G., Maurer, P. C., Yao, N. Y., et al. (2013, July 31). Nanometre-scale thermometry in a living cell. Nature, 500(7460), 54–58. https://doi.org/10.1038/nature12373
In text: (Kucsko et al., 2013)
Also returned: MLA 9, Chicago 17 with footnote, Harvard, plus BibTeX and RIS.

Journal metadata comes from Crossref, so volume, issue and pages are the publisher's own.

How this compares

Where another tool is the better choice, it says so.

ZoteroBib

BETTER THEREFrom the Zotero team, free, no ads, and the best metadata extraction on the web — it uses the same translators as Zotero itself, which beats reading meta tags.

BETTER HEREIt shows one style at a time and does not tell you what it failed to find. Here all four styles appear together with the in-text forms and an explicit list of missing fields.

Cite This For Me / MyBib

BETTER THEREEnormous style libraries — hundreds of variants including institution-specific ones this tool does not cover.

BETTER HEREBoth are ad-supported and push a subscription, and both invent plausible-looking references when metadata is thin. This has no ads, no upsell, and says n.d. rather than guessing a year.

Zotero itself

BETTER THEREThe right answer for a thesis: it stores sources permanently, syncs, inserts citations in Word and reformats a whole bibliography when you change style.

BETTER HEREIt needs installing, an account for sync, and a browser connector. For five sources tonight, pasting a link here is faster and there is nothing to set up.

Your questions, answered.

Does APA 7 still use “Retrieved from”?: No. Give the bare URL with no lead-in and no final full stop. A retrieval date appears only for pages designed to change, such as a wiki or a live feed.
What if the web page has no author?: The title moves into the author position and the date follows it. Check first whether an organisation is the author — on institutional pages it almost always is.
Is it free?: Yes, and it stays free: no credits, no account, no ads, no watermark, and the BibTeX and RIS downloads cost nothing either. Fifty sources a day is the only limit, and it exists to stop scripts, not students.
Which styles are supported?: APA 7th, MLA 9th, Chicago 17th (both the note and the author-date forms) and Harvard as set out in Cite Them Right. All four are generated from one lookup.
Can I paste a DOI or an ISBN instead of a link?: Yes, and you should when you have one. A DOI goes to Crossref and comes back with the journal, volume, issue and pages already correct — far better metadata than any article page publishes.
Why is the author missing or wrong?: Because that is what the page published. Many sites list no author, or list a staff account. The output says when no author was found; replace a staff byline with the person named on the page, or with the organisation.
Why does my reference say n.d.?: No publication date was found in the page's metadata. Check the footer, the byline and any “last updated” line, and add it by hand — n.d. is correct only if the source genuinely carries no date.
Does it fix the capitalisation of titles for APA?: No, on purpose. Re-casing a title automatically also lower-cases proper nouns, which introduces errors that look like yours. The title comes through as published and you adjust it where you can see the names.
Can it cite a book chapter or an edited collection?: Yes — set “Source is a” to book chapter. Detection cannot tell a chapter from a book from metadata alone, and the styles treat them very differently.
Can I look a book up by its title?: Yes, if you do not have the ISBN: type the title and the closest match in Open Library is used. It is a guess, so the result names what it matched and asks you to check it — the ISBN from the back cover is always more reliable.
What about a paywalled or login-only page?: Nothing here can see behind a paywall, so the lookup fails and says so. Cite it from what you can see, or find the article's DOI, which works regardless of access.
Is this accurate enough to submit?: The formatting is; the metadata is only as good as what the source published. Read the reference once against the page — ten seconds — and fix the author or date the warnings flag.
Do you store what I cite?: The citation stays attached to your job so you can come back to it, and nothing is published or shared. No account, no email.

APA Citation for a Website

Add your source.