Is the Word Cloud Generator free?

Yes, Word Cloud Generator is totally free :)

Can I use the Word Cloud Generator offline?

Yes, you can install the webapp as PWA.

Is it safe to use Word Cloud Generator?

Yes, any data related to Word Cloud Generator only stored in your browser (if storage required). You can simply clear browser cache to clear all the stored data. We do not store any data on server.

How does the Word Cloud Generator work?

Your text is split into tokens on word boundaries, lowercased unless you ask for case sensitivity, stripped of punctuation, links and stop words, and then counted into a frequency table. The top terms are scaled to a font size proportional to their count and packed along an outward spiral so no two words overlap. Everything runs in your browser, so the text you paste never leaves your device.

What are stop words and should I remove them?

Stop words are high-frequency grammar words such as the, and, of, to and is that appear in every English document regardless of topic. Leaving them in produces a cloud dominated by words that say nothing about the subject, which is why removal is on by default. Turn it off only when the function words themselves are what you are studying, such as in stylometry or language teaching.

What is the difference between frequency and keyword density?

Frequency in the table is a term's share of the words that survived filtering, so it answers the question of which topics dominate the filtered vocabulary. Keyword density in SEO mode divides by every word in the document, including stop words, because that is the denominator search marketers and content briefs use. The same term therefore shows a higher frequency than density.

Why do some words disappear from the cloud?

A term can be dropped for four reasons: it is shorter than the minimum word length, it appears fewer times than the minimum frequency, it matched the stop-word or custom exclusion list, or it ranked below your maximum word count. A fifth case is layout — if the canvas fills up before every word is placed, the omitted count appears under the cloud and a larger canvas or fewer words fixes it.

Can word clouds be misleading?

Yes, and the tool shows a bar chart alongside the cloud for exactly that reason. Font size is judged by area, not height, so readers systematically over-read large words, and long words look more important than short ones at the same size. A cloud also measures repetition rather than importance, so a term mentioned once in a decisive conclusion is drawn smaller than filler repeated throughout.

How do I export the cloud for a slide or a report?

Download SVG for a vector file that stays sharp at any size, which is the right choice for slides and posters. Download PNG when you need a raster image, choosing 2x or 3x for retina screens and print. The frequency table exports separately as CSV for spreadsheets, as a Markdown table for documentation, or as a plain text summary that includes every statistic.

Word Cloud Generator - Free Text Frequency Tool

Rank	Count	Frequency
1	4	9.09%
2	3	6.82%
3	3	6.82%
4	2	4.55%
5	2	4.55%
6	2	4.55%
7	2	4.55%
8	1	2.27%
9	1	2.27%
10	1	2.27%
11	1	2.27%
12	1	2.27%
13	1	2.27%
14	1	2.27%
15	1	2.27%
16	1	2.27%
17	1	2.27%
18	1	2.27%
19	1	2.27%
20	1	2.27%
21	1	2.27%
22	1	2.27%
23	1	2.27%
24	1	2.27%
25	1	2.27%
26	1	2.27%
27	1	2.27%
28	1	2.27%
29	1	2.27%
30	1	2.27%
31	1	2.27%
32	1	2.27%
33	1	2.27%

Word Cloud Generator: Turn Any Text Into a Frequency Picture

A word cloud generator reads a block of text and draws each word at a size proportional to how often it appears, so the themes of a long document become visible in a single glance. Paste an essay, an interview transcript, a pile of product reviews, a survey export or a column of CSV values, and this tool tokenises the text, removes filler words, counts what is left and packs the top terms into a shareable image. Alongside the picture it produces the numbers behind it: a ranked word frequency table with counts, percentages and proportional bars.

How the Tool Counts Words

Counting sounds trivial and is not. The text is first cleaned of HTML tags and, optionally, of links and email addresses that would otherwise contribute meaningless tokens. It is then split on Unicode word boundaries, so accented letters survive and internal apostrophes and hyphens keep don't and state-of-the-art intact. Each token is trimmed of stray punctuation, lowercased unless case sensitivity is on, and tested against the filters: minimum length, minimum frequency, numeric tokens, the built-in stop-word list and any custom exclusions you add.

The survivors are counted into a map and ranked, with alphabetical tie-breaking so equal counts always appear in the same order. Frequency is then count ÷ counted words × 100. Two derived measures sit beside it: lexical diversity, the ratio of unique terms to counted terms, which rises with richer vocabulary and falls with repetition, and average term length in characters.

Stop Words, Stemming and N-Grams

Stop words are the grammatical glue of a language — the, and, of, to, is — and they top the frequency list of every English document regardless of subject. Removing them is on by default because a cloud full of function words tells you nothing. Turn removal off when the function words themselves are the object of study, as in authorship analysis or language teaching.

Plural and stem grouping merges variants such as test, tests and testing into one entry so a single idea is not split across three small words. Switching the n-gram size to two or three counts adjacent word pairs or triples instead, which surfaces multi-word themes that single words scatter: battery life and customer support are visible as phrases but invisible as isolated tokens. Phrases never span a full stop or a line break, so no nonsense pairs are created across sentence boundaries.

Sizing and Layout

Font size is mapped from the count with your chosen curve. A linear scale is faithful to the raw numbers but lets a single dominant word flatten everything else; square root compresses the top end so mid-frequency words stay readable; and a logarithmic scale is the flattest option, useful when counts span several orders of magnitude. Words are then placed largest-first along an outward Archimedean spiral, and each candidate position is rejected if its bounding box collides with an already-placed word or leaves the shape mask. Anything that never finds a gap is reported rather than drawn on top of its neighbours.

Reading a cloud honestly

Font size is judged by area rather than height, so large words are over-read, and long words look more important than short ones drawn at the same size. Use the bar chart and the ranked table for the precise comparison that a cloud deliberately blurs.

Keyword Density for SEO

SEO mode reports keyword density for one-, two- and three-word terms at once, dividing by every word in the document rather than by the filtered vocabulary, because that is the denominator content briefs use. Any term above your threshold — three percent is a common starting point — is flagged as potential over-optimisation. Density is also expressed per thousand words for teams that report it that way. Use it to check that a page actually mentions its target term, and to catch the repetition that reads as keyword stuffing.

Comparing Two Documents

Comparison mode measures each term's share of the vocabulary in two texts and reports the difference in percentage points, sorted by the largest shift. Point it at last year's reviews versus this year's, two competitors' landing pages, or the transcripts of two focus groups, and the terms that grew or shrank appear at the top. Because the comparison uses relative frequency rather than raw counts, the two texts do not need to be the same length.

Exports and Privacy

Download the cloud as SVG for slides and posters where it must stay sharp at any size, or as PNG at one, two or three times resolution for screens and print. The frequency data exports separately as CSV for spreadsheets, as a Markdown table for documentation, or as a plain text summary containing every statistic. Every step — tokenising, counting, layout and rasterising — happens in your browser, so confidential feedback and unpublished drafts never leave your machine.

Word Cloud Generator

About This Tool