Accessibility statement

Our commitment

Free Tools is meant to be usable by everyone, including people who use a keyboard only, a screen reader, a screen magnifier, voice control, high-contrast or forced-colour modes, or who prefer reduced motion. We treat an accessibility barrier as a bug and fix it.

Conformance status

The 1,881 pages of this site (the tool hub, the About page, this page and 1,878 tool pages) conform to the Web Content Accessibility Guidelines (WCAG) 2.2 at level AA, for the scope and with the exceptions described below, notably the content of previews and files that tools generate from what you type. Automated checks covered every page; manual keyboard and zoom checks covered a sample of 125 pages and screen-reader testing covered three pages, as set out under How it was tested. This is our own assessment, made on 23 October 2026; it is not a certification by a third party.

WCAG 2.2 AA includes every requirement of WCAG 2.1 AA, which is what clause 9 (web content) of the European standard EN 301 549 V3.2.1 refers to. The site is therefore also aligned with that standard. The European Accessibility Act (Directive (EU) 2019/882) sets its own scope of products and services; we use the same standard as our yardstick whether or not the Act applies to this free site.

What was tested

Pages
All 1,881 pages: the hub, About, this statement and every one of the 1,878 tool pages.
Configurations
Light theme and dark theme (both the site toggle and the system colour-scheme setting), each at a 1280 px wide viewport and at a 320 px wide viewport (the equivalent of 400% zoom): four configurations per page, 7,524 page audits.
States
Each tool page was audited after its sample had been run and after waiting for late renders. Tools with options were audited again with one non-default option set. The 146 tools that draw their own interface (including camera, microphone, location, audio, video, 3D viewer, in-browser AI, speech-to-text, chat, chart, map, quiz, piano, drum machine, whiteboard, geometry, puzzle, font-maker and timer tools) were also audited after clicking their buttons and changing their controls; where a sample image, PDF or audio file could be supplied it was loaded first.
Browser
Google Chrome 151 on Linux. Other browsers were not tested.
Date
23 October 2026.

How it was tested

  • Automated: axe-core 4.14.0 with the rule sets wcag2a, wcag2aa, wcag21a, wcag21aa, wcag22aa, best-practice, EN-301-549 and ACT (the AAA-only enhanced-contrast rule was switched off). Result: 0 violations in the page itself in every page, theme and viewport. Items axe reports as needing review were triaged by hand (see below).
  • Scripted checks on every page: keyboard focusability and visible focus ring, controls handled only by mouse events, control size of at least 24 by 24 CSS pixels (2.5.8), unnamed canvases, positive tabindex, running animations under prefers-reduced-motion, and horizontal scrolling at 320 px.
  • Manual keyboard-only walk-throughs of 125 pages: text tools, two-input tools, option-heavy tools, file tools (including PDF, document, image and database tools), the hub search, the theme toggle, the FAQ disclosures, the skip link, timers, games, board and drawing tools, microphone, recorder and tuner tools, in-browser AI tools, location and display tools, a 3D model viewer, piano and drum-machine tools, map and quiz tools, video and PDF editors, chart tools, and tools that draw on a canvas. We checked that every function can be operated, that there is no keyboard trap, that the focus order is logical, that focus stays visible and is not hidden, and that focus is not lost after Run, Reset or the theme toggle.
  • Text spacing (WCAG 1.4.12) with the standard override style sheet on 60 pages, reflow at 320 px on all pages, forced-colours emulation (Windows High Contrast) on 28 pages, and reduced-motion emulation on all pages.
  • Colour contrast of every design token pair in both themes, including control borders and focus rings (at least 3:1) and text (at least 4.5:1). Primary buttons are dark text on yellow (13:1) in both themes; the focus ring is near-black on light and yellow on dark (at least 12:1 against the page).
  • Accessibility tree inspection of the custom tools (names, roles and states) and one screen reader session: Orca 46.1 with Chrome on Linux, on three pages (JSON Formatter, Minesweeper, Kanban Board); the other tools were checked through the accessibility tree and the keyboard, not with Orca. We checked that page structure, field and button names, toggle states, game cells and the 'Result updated' status message are announced. No other screen reader (NVDA, JAWS, VoiceOver, TalkBack) and no voice-control software has been used.

How the site is built for accessibility

  • Every page has one h1, a logical heading order, named landmarks, a skip link and a unique title; the language is declared.
  • Every form control has a label, and tool errors are announced and say what went wrong and where. Results are not read out automatically on every keystroke; pressing Run or Reset announces a short status instead.
  • Both themes keep text at 4.5:1 or better. Content reflows down to 320 px without two-dimensional scrolling; long outputs wrap, and any region that scrolls can be reached with the keyboard.
  • Animation respects the reduced-motion setting. Animated previews can be paused, and the Screen Flashlight tool starts at 2 flashes per second.
  • Things you would normally do by dragging or clicking on a picture have a keyboard alternative: numeric box fields for Image Redaction, position and pixel fields for Screen Ruler and Colour Picker from Image, a Type mode for Signature Generator, and buttons for moving cards on the Kanban Board.

Known limitations

  • Previews produced by tools are shown in sandboxed frames (for example rendered HTML, SVG, CSS, email, invoice and banner previews). We give each frame a title and the preview document a language and title, but the preview shows what you typed: its content, colours and structure are yours, and some samples are deliberately not perfect. Known cases: the colours picked in the Shadcn Theme Generator and Social Preview samples, a scrollable demo inside the CSS Scrollbar Generator and Graphviz DOT preview, and disabled task-list checkboxes that Markdown Preview and the Markdown Cheatsheet produce without a label.
  • Documents and files that tools generate (PDF, SVG, HTML, images, audio, video, WAV, ZIP) are not guaranteed to be accessible. Invoice, receipt and resume documents are plain HTML printed by the browser and are not tagged PDFs.
  • Tools whose purpose is visual or physical depend on sight or on a pointer: drawing a signature (use Type mode instead), image redaction, the colour-blindness simulator, image diff, the screen ruler, the decision wheel (the result is also written as text), games, the compass and level sensors, and the click-speed test (it can also be run with Enter or Space, but speed is the point of the test).
  • Screen Flashlight & Beacon has a strobe mode that can be raised to 10 flashes per second, above the 3 per second general flash threshold. It is off by default, carries a warning, and is capped at 3 per second when your system asks for reduced motion.
  • Audio and video recorders (Screen Recorder, Webcam Recorder, Tone Generator, Noise Generator, Metronome) record or play what you choose. Recordings have no captions or transcript, because they are your own material. The controls and status messages are accessible.
  • Live displays (timers, clocks, meters, sensors) update continuously because that is their function. The World Clock has a Pause updates button; timers have Pause or Stop.
  • Tools that show data in a known format (regular expressions, code, hashes, binary, protocol dumps) present it as plain text in a scrollable region; some of it is dense and has no simpler wording.
  • The video tools (trimmer, converter, compressor and the others) were audited and keyboard-tested only up to the point where a video file is needed, because no sample video was available in the audit, so their processing and result states were not exercised. Some tools show a preview of colours you chose or standard terminal colours (ANSI Log Color Renderer, BBCode to HTML, Code Syntax Highlighter) that do not reach 4.5:1 by design. The Teleprompter Recorder clips its text to a scrolling window on purpose. The Drum Machine and Step Sequencer expose every step as its own tab stop, which makes keyboard use slow. The map tools (Travel Map Pinner, Geography Quiz Map) can be clicked, and each also has a keyboard route (a checklist and a typed answer). The Email Signature Generator and Avatar Generator previews contain sample links and text that axe flags at a 320 px viewport (links below 24 px high, text over a generated image). The AI tools (fill-mask, question answering, language identification, sentiment, speech to text, offline chat) load a language model first. Their status and error messages are marked up to be announced and the tools can be operated by keyboard, but we could not run the model download itself in the audit, so the real loading states were not tested. Tools that need the microphone, camera or location only work after you grant the browser permission, and the permission prompt is the browser's own. The Code Playground runs your own HTML, CSS and JavaScript in a sandbox; what it renders is not checked.
  • Windows High Contrast and reduced motion were tested in Chrome emulation, not on a physical Windows machine.

Items automated tools could not decide

axe-core reports a small number of items as needing review. We checked them by hand and found no failure: text drawn over elements it cannot analyse (the ASCII art result and the compass needle), text areas whose resize handle overlaps them, buttons whose visible label is an emoji or symbol with a word as the accessible name (Emoji Picker, Unicode Character Map, grid editor buttons), and table cells that contain only a number.

What this statement does not cover

  • Third-party websites you reach through links, and the files you load into tools.
  • Browsers, operating systems and assistive technologies other than those listed above.
  • Future changes: tools are added and edited. The statement is reviewed whenever the audit is repeated and the date at the top is updated then.

Feedback and contact

If you run into a barrier, tell us the page address, the tool you were using, your browser and any assistive technology, and what you expected to happen. We aim to reply within 14 days.

Contact details will be published here.

Enforcement procedure

If you are not satisfied with our reply, you can complain to the national body responsible for enforcing digital accessibility requirements in the country of the site operator.

[TODO for the site owner: name the operator and its country, and give the name, website and contact details of the national enforcement body or ombudsman for that country (for Member States, see the national authorities listed under Directive (EU) 2019/882 and Directive (EU) 2016/2102). Remove this note when done.]

Technical specification

Accessibility of this site relies on HTML, WAI-ARIA, CSS and JavaScript. All of them are supported by the browsers used for testing.