Frequently Asked Questions

Yes. Every one of our free online text tools is free forever — no sign-up, no paywall, no watermarks on outputs, and no trial limits.

No. Textbench is a static website — there is no backend that receives anything you paste or upload. Every tool runs entirely as JavaScript inside your browser tab. You can verify this by opening DevTools' Network tab, or by disconnecting from the internet after the page loads.

No account is required. Open any tool, paste your text or drop a file, and you'll see results immediately. Nothing about you is collected beyond standard anonymous Google Analytics pageviews.

Yes. Every tool is responsive and works on mobile, tablet, and desktop browsers. No app install is required — bookmark the page or add it to your home screen.

Once a tool page has loaded once, the actual processing runs in your browser, so most tools will keep working even if you go offline. Pages have to be reachable for the first load, but no further server contact is needed.

Any modern browser released in the last few years — Chrome, Edge, Firefox, Safari, Brave, Opera, and their mobile equivalents. Some tools rely on the Web Crypto and Clipboard APIs which are standard in all of the above.

The text never leaves your browser — that's the central design promise. See the privacy page for the full statement. That said, as a general rule, avoid pasting production secrets into any third-party tool you can't audit yourself.

Standard anonymous Google Analytics pageview data — referrer, page path, country, device class. We never collect or transmit the contents of what you paste into a tool.

Google Analytics sets its own cookies. The site itself stores only a couple of UI preferences (theme, the 'remove duplicates' toggle on the Pattern Extractor) in your browser's localStorage. None of that is transmitted off your device.

Open the Pattern Extractor, paste your text, and make sure the "URLs" pattern chip is selected (it is by default). The extractor finds http, https, and bare domains. You can then copy the list, export to TXT/CSV/JSON, or click "Open all" to open every URL in a new tab.

Paste the text (or drop the .txt / .csv file) into the Pattern Extractor and tap the "Emails" pattern chip. The tool detects standard RFC-ish email syntax — you can dedupe the results and download them as CSV for spreadsheet use.

Yes. After extraction, the "Open N links" button opens every URL — and the "Batch…" dropdown opens them in groups of 5, 10, or 20 with a 5-second delay between batches, so your pop-up blocker doesn't kill the whole list.

URLs, email addresses, phone numbers (international + extensions), FTP URLs, IPv4, IPv6, dates, times, hex colors, Twitter handles, hashtags, Markdown links, and ISBN numbers — 13 patterns total. Toggle any combination as chips above the textarea.

Yes — the Export menu next to the matches list downloads either plain text (one match per line), CSV with type and value columns, or JSON. Filenames include a timestamp so repeated runs don't overwrite each other.

Paste your list into Deduplicate Lines. The first occurrence of each line is kept, the rest are dropped, and the original order is preserved. You can also ignore case, trim whitespace, and skip empty lines.

Use the Sort Lines tool. It sorts case-insensitively by default and uses natural ordering, so "item2" appears before "item10" instead of after it. You can also sort numerically, by line length, or reverse the order.

That's called "lexicographic" sort, where digits are compared as characters — so '1' beats '2'. The Sort Lines tool uses natural ordering by default, which compares the embedded numbers numerically: item2 → item10 → item100.

Yes. Find & Replace has a regex toggle, plus case-sensitivity and whole-word options. Capture groups in your regex are available as $1, $2, … in the replacement string.

Paste the original on the left and the changed version on the right in Text Diff. The tool highlights added lines in green and removed lines in red, with line numbers on each side and counts of how many lines were added, removed, and unchanged.

Line-by-line, using a longest-common-subsequence algorithm — the same approach Git uses for its unified diff. It handles moved blocks reasonably well but is not a word-level diff.

In the Word & Character Counter, anything separated by whitespace counts as a word after leading and trailing punctuation is stripped. So "don't" is one word, but "hello," is also one word because the comma is trimmed.

Reading time uses ~238 words per minute (the median for adult silent reading on screens) and speaking time uses ~140 WPM (typical conversational pace). Both numbers come from published studies on reading and speech rates.

The Case Converter handles all of those, plus PascalCase, CONSTANT_CASE, Title Case, Sentence case, and a case-invert mode. Programmer cases split your input on whitespace, punctuation, and case boundaries.

Title Case capitalises the first letter of every word ("Textbench Is Fast"). Sentence case capitalises only the first letter of each sentence ("Textbench is fast."). The Case Converter supports both.

Paste your JSON into the JSON Formatter. Pretty-print mode indents with 2 or 4 spaces; minify mode collapses to a single line. Invalid JSON shows the parser error inline so you can see exactly where the problem is.

Use the Base64 Encode / Decode tool. Paste any text — full UTF-8 is supported (emoji, accented characters, etc.) — and copy or download the encoded result. Toggle "URL-safe" to swap +/ for -_ and drop padding, the variant used in JWTs.

No. Base64 is an encoding, not encryption — anyone who can read the Base64 string can decode it back to the original. Use Base64 to safely transport binary in text contexts (JSON, HTTP headers, data URIs), not to hide secrets.

Any time a value contains characters that would change a URL's meaning — spaces, &, =, ?, non-ASCII, or emoji. The URL Encode / Decode tool uses encodeURIComponent semantics by default and offers a form-encoding mode (spaces as +) for application/x-www-form-urlencoded bodies.

MD5 produces a 128-bit hash, SHA-256 produces a 256-bit hash. Cryptographically, MD5 is broken (collisions can be constructed in seconds) and shouldn't be used for security; SHA-256 is the modern default for signing, integrity checks, and password hashing (with a slow KDF like Argon2 on top).

For non-security use cases — content fingerprinting, cache keys, deduplication — MD5 is fine and fast. For anything where collisions matter (signatures, tamper detection, password storage), use SHA-256 or stronger. The Hash Generator computes both side by side.

No — cryptographic hashes are one-way functions. The closest thing is a rainbow-table lookup, which can recover short, common inputs (single English words, short numbers) but fails on anything reasonably long or random. Our hash generator computes hashes, but no online tool can truly "decrypt" one.

Open the QR Code Generator, paste your URL or text, pick a size, and download as PNG or SVG. There are no watermarks, no sign-up, and the codes generate entirely in your browser. You can also tune the error-correction level for printing at small sizes.

A QR code can technically encode up to ~4,296 alphanumeric characters or ~2,953 bytes, but anything beyond a few hundred characters becomes hard to scan. For URLs longer than ~100 characters, use a URL shortener and put the short URL in the QR code instead.

Lorem Ipsum is scrambled Latin placeholder text dating to the 1500s. It mimics the rhythm and letter distribution of real prose without being legible, so reviewers focus on layout rather than the words.

Up to 500 of whichever unit you pick — paragraphs, sentences, or words — on the Lorem Ipsum Generator. You can toggle whether the first paragraph starts with the classic "Lorem ipsum dolor sit amet…" opening.

Paste any colour value into the Color Converter — HEX (#RRGGBB or #RRGGBBAA), rgb(…), or hsl(…). The other two formats update live in the other fields, and the preview shows transparency over a checkerboard.

HSL (hue, saturation, lightness) is much friendlier for palette design. You can keep hue and saturation fixed and just nudge lightness to get a tonal range, or rotate hue for accent colours that still feel cohesive. RGB is great for picking exact colours but bad for shifting along these axes.