Skip to content
Lizely

Word Cloud Generator

Turn pasted text into a downloadable word cloud without uploading your writing.

Privacy: your files never leave your device. All processing happens locally in your browser.

How to use

  1. 1.Paste or type the text you want to visualize in the Text field.
  2. 2.Optionally enter words to exclude and choose whether capitalization should remain separate.
  3. 3.Select Generate word cloud, review the preview, and download the full-size PNG.

About Word Cloud Generator

Word Cloud Generator turns an article, speech, lesson, interview transcript, survey response, meeting note, or brainstorming list into a compact visual summary. Paste text into the editor and select Generate word cloud. Words that occur more often receive larger type, while less frequent terms appear smaller. The result can help you spot recurring vocabulary, prepare a presentation graphic, compare the emphasis in a draft, or create a visual prompt for a workshop. When the cloud is ready, download it as a 1200 by 800 pixel PNG that is suitable for slides, documents, classroom materials, and social posts. The exported image uses its full natural pixel dimensions rather than a reduced screenshot of the on-page preview.

All counting, layout, drawing, and PNG generation take place in the current browser. Your text is not uploaded to Lizely, sent to an analysis service, or saved to an account. This local workflow is useful for unpublished drafts, internal notes, student writing, and other material that you do not want to submit to a remote processor. The tool does not perform language detection, semantic analysis, sentiment analysis, or topic modeling. It simply extracts letter-and-number word tokens, counts them, assigns relative font sizes, and places them on a canvas. The same input and settings produce the same arrangement, so a cloud will not jump unpredictably between renders.

Token matching is Unicode aware, allowing words written with many non-English scripts to be counted instead of restricting input to ASCII letters. Numbers can also appear as tokens. Common punctuation separates words, while an internal apostrophe or hyphen can remain part of a term such as don't or co-op. By default, uppercase and lowercase forms are combined, so Cloud, cloud, and CLOUD contribute to one count. Enable the case-sensitive option when capitalization itself matters. Compatibility normalization is applied before counting so visually compatible forms are treated consistently. These practical rules are designed for readable summaries, not linguistic research, and they may not match every language's rules for compound words or word boundaries.

No built-in stop-word dictionary is included. A universal list would silently make assumptions about language, context, and what the author considers unimportant. Instead, enter your own excluded words in the optional field. For an English article you might exclude the and and; for a product review you may choose to exclude the product name; for survey feedback you may leave every term visible. Exclusions use the same tokenization and case setting as the main text. Review the visible cloud after changing the exclusion list and generate it again to apply the updated settings.

The layout starts with the highest-frequency terms and follows a deterministic spiral from the center. Each candidate is measured with the browser's real canvas text metrics before placement. The tool checks canvas boundaries and existing rectangles, then skips a word when it cannot find a collision-free position. This may happen with an extremely long token, a large number of unique words, or a crowded cloud. Up to 80 of the most frequent terms are considered so rendering remains responsive. A message reports how many words were placed and tells you when some terms did not fit. To make more room, remove low-value text, add exclusions, shorten unusually long tokens, or focus on one section at a time.

Word frequency is an exploratory signal, not a judgment of meaning or importance. Repeated boilerplate, quotations, navigation labels, names, and copied footers can dominate a cloud even when they are not the central idea. Clean the source text and choose exclusions deliberately before using the image in a report. The generator does not identify misinformation, sensitive data, plagiarism, authorship, quality, or intent. Read the original material and use human judgment for any consequential interpretation. For a precise numerical audit, use a word-count or text-analysis workflow alongside the visual cloud rather than estimating exact counts from font sizes.

Methodology & sources

The browser normalizes text, extracts Unicode letter-and-number tokens, applies the visitor's exclusions and case preference, counts occurrences, scales font sizes by square-root frequency, measures real canvas text, and places terms along a deterministic spiral while rejecting collisions and out-of-bounds rectangles. The full-size 1200 by 800 canvas is the PNG export source.

Frequently asked questions

Does the tool upload or store my text?
No. Token counting, layout, canvas drawing, and PNG creation all happen locally in your browser.
Why are common words still visible?
The generator deliberately has no built-in stop-word dictionary. Enter any terms you want removed in the Exclude words field.
Why did some words not appear in the cloud?
The tool considers up to 80 frequent terms and skips a word if its measured rectangle cannot fit without crossing the canvas boundary or another word.

Related tools

Text Tools guides

View all