Frequently Asked Questions

CopyRefine is a set of three free writing tools: a Filler Word Detector, a Readability Score calculator, and a Tone Detector. All tools run entirely in your browser — no AI, no server uploads, no signup required. They help writers, marketers, and anyone who writes in English polish their copy and communicate more clearly.

No. CopyRefine requires no account, no signup, and no email. All tools are free to use immediately with zero friction.

No. CopyRefine does not use AI, machine learning, or any black-box algorithms. Every tool is powered by transparent rules — word lists that you can inspect, mathematical formulas (like the Flesch Reading Ease equation), and simple keyword matching. There is no LLM, no neural network, and no external API call involved.

Yes, all three tools are completely free. There are no paid plans, no usage limits, and no hidden charges. If that ever changes, we will clearly announce it.

Since everything runs locally in your browser, you can use CopyRefine offline as long as you have loaded the page at least once while online. The JavaScript and word libraries are cached by your browser.

Currently, all three tools are English-only. The filler word library, formal/friendly word libraries, and syllable counting algorithm are all designed for English text. We may explore other languages in the future.

CopyRefine works on all modern browsers: Chrome, Firefox, Safari, and Edge. Internet Explorer is not supported. For the best experience, use the latest version of your preferred browser.

The Filler Word Detector uses a library of over 200 common filler words and phrases, including "actually", "basically", "very", "really", "just", "sort of", "kind of", "literally", "at the end of the day", and many more. The library focuses on English filler words commonly used in business and marketing writing.

No. The tool highlights potential filler words for your review. Whether to remove them is your decision. Some filler words serve legitimate purposes — they can set a conversational tone, add emphasis, or create rhythm. The tool gives you the information; you make the call.

Compression rate shows the percentage of words that could be removed if all detected filler words were deleted. For example, a compression rate of 20% means roughly one in five words is a potential filler. This is a theoretical maximum — in practice, you would keep some of them.

Not through the interface. However, since the word library is a plain JavaScript file (data/filler-words.js), you can inspect it directly in your browser's developer tools. The library is transparent and accessible.

The Flesch Reading Ease formula is a readability metric developed by Rudolf Flesch in 1948. It calculates a score from 0 to 100 based on two factors: average sentence length and average syllables per word. Higher scores indicate easier readability. The formula is: 206.835 minus 1.015 times (words/sentences) minus 84.6 times (syllables/words).

For general web content, aim for a score of 60-70. This is considered "Standard" and is appropriate for readers aged 13-15. For technical documentation, 30-50 (College level) may be acceptable. For marketing copy aimed at a broad audience, 65-80 works well. For children's content, target 80-100.

The syllable counting uses a heuristic algorithm (vowel-group detection) with a database of common exceptions. It is not 100% accurate — no automated syllable counter is — but it is accurate enough for meaningful readability estimation. The error margin typically affects the final Flesch score by no more than 3-5 points.

No. The Flesch Reading Ease score is a well-established heuristic that has been used for decades in education, publishing, and government. However, it is a simplified formula that measures only sentence length and word complexity. It does not account for organization, visual design, reader motivation, or subject matter expertise. Use it as a helpful guide, not a definitive assessment.

No. The Tone Detector uses keyword matching and cannot understand context, sarcasm, irony, humor, or nuance. For example, "Great, another software update" would be flagged as containing the friendly word "great" even though the tone is sarcastic. Always use your own judgment alongside the tool's output.

For most business emails, a "friendly but professional" tone works best. This means using "you", "we", and "thanks" to create warmth, while maintaining clear structure and respect. Avoid overly formal language like "pursuant to" or "heretofore" unless you are writing legal correspondence.

Yes! The formal and friendly word libraries are open and transparent. You can inspect them directly in your browser's developer tools or view the source files at data/formal-words.js and data/friendly-words.js. We believe in complete transparency — no black boxes.

No. All tool analysis happens entirely in your browser using JavaScript. No text you paste into any CopyRefine tool is transmitted to any server, stored, logged, or shared. This is a core design principle of the site.

CopyRefine uses exactly one cookie: copyrefine_cookie_consent. This stores whether you accepted or declined the cookie consent notice. It contains only the values "accepted" or "declined" and expires after 365 days. No tracking cookies, analytics cookies, or advertising cookies are used.

No. CopyRefine does not use Google Analytics, Facebook Pixel, or any other third-party tracking scripts. The site is designed to be privacy-first. If this changes in the future, we will update the privacy policy and cookie notice accordingly.

Yes. CopyRefine follows GDPR principles: we collect no personal data, we use only one essential (non-tracking) cookie with your consent, all processing is local, and we provide transparent information about our data practices in our privacy policy. Since we collect no data, there is nothing to access, rectify, or erase — but if you have questions, contact us at jinli156125@163.com.