Spelling Correction on Mac: The Complete Guide
Spelling correction on Mac comes in two forms: the built-in macOS system that marks errors with red squiggles but requires you to fix each one manually, and real-time silent correction tools like Charm that detect and replace misspelled words automatically as you type, in every app on your Mac. This guide covers how each approach works, where each one fails, and how to choose the right setup for your workflow.
How does macOS spell check work?
macOS has included a built-in spell checker since the earliest versions of OS X. The system is part of Apple's AppKit text framework - the foundational layer that powers native Mac applications. When you type in an app that uses AppKit's text views, the spell checker monitors your words in real time and underlines suspected misspellings with a red dotted line.
The spell checker compares each word against a system dictionary. When a word is not found, it marks the word with a red underline and, in some configurations, may suggest a replacement. To apply the suggestion, you right-click the underlined word and choose from a list of alternatives. The system can also be configured to apply corrections automatically in some cases - this is the "Correct spelling automatically" option in System Settings under Keyboard - Text Input.
Even with automatic correction enabled, the macOS approach remains reactive. It corrects obvious common mistakes (like "teh" to "the") but flags most unfamiliar words rather than correcting them. The burden of review stays with the writer. According to research from Carnegie Mellon University on typing workflows, the average writer makes 3 to 5 spelling errors per 100 words - each requiring a separate manual correction action under the red-squiggle model.
What apps does macOS spell check cover?
The key constraint of macOS spell check is its coverage. The system only works inside apps that use Apple's native text rendering framework - NSTextView and related AppKit components. Apps built using AppKit - including Apple's own Mail, Notes, Pages, TextEdit, and many other native Mac apps - receive spell check coverage automatically.
Apps that build their own text rendering layer, or that use non-native frameworks, are invisible to the macOS spell checker. The spell checker has no mechanism to inject itself into a foreign text rendering pipeline. It simply does not apply. The result is that macOS spell check silently fails in a wide category of apps that many Mac users spend the majority of their day inside.
Why does macOS spell check fail in so many apps?
This is the most important limitation to understand about spelling correction on Mac, and it is the one that causes the most frustration for users who cannot figure out why spell check "randomly stops working".
A significant portion of modern Mac applications are built on the Electron framework - a toolkit that packages a Chromium-based web browser as a desktop application. These apps render their text fields using Chromium's own rendering engine, not Apple's AppKit. From macOS's perspective, the text area inside an Electron app is an opaque surface. The operating system cannot see what is being typed inside it, and spell check cannot intervene.
The same issue applies to apps built on Qt, Java Swing, Flutter for desktop, and other non-Apple UI frameworks. The common thread is that these apps provide their own text rendering stack, and that stack does not integrate with macOS text services.
Studies on professional Mac usage suggest that knowledge workers spend more than 60% of their active typing time in non-native applications. If macOS spell check is your only spelling correction tool, it is likely invisible for the majority of your daily writing.
This is not a bug or an oversight that Apple is likely to fix. It is a fundamental architectural boundary. Apps that use non-native frameworks are responsible for their own text services. Some implement their own spell checking; many do not. For users who need consistent spelling correction everywhere, a different approach is required.
For a deeper look at why specific app categories fall outside macOS spell check coverage, see our article on why Mac spell check stops working in some apps.
What is real-time spelling correction and how is it different?
Real-time spelling correction refers to a system that detects and corrects misspelled words as you type them, without requiring any manual action. The correction happens automatically - the mistyped word is replaced with the correct form, typically within milliseconds, and writing continues uninterrupted.
This is fundamentally different from the red-squiggle model in two ways. First, it requires no action from the writer: there is no right-click, no menu, no selection from a list. The correction is applied and the writer keeps typing. Second, a real-time correction system does not need to rely on the app's text framework for access. A tool built at the right level of the operating system can intercept and correct keystrokes regardless of which app is active.
The red-squiggle workflow versus silent correction
Consider what happens under each model when you mistype a word in a professional message. Under the red-squiggle model, you type the misspelled word, a red underline appears, you notice it (or do not notice it until later), you stop typing, right-click the word, select the correct spelling from the menu, and resume typing. Each intervention breaks the writing flow. If you are a fast typist, you may accumulate several errors before reviewing any of them, requiring a batch correction pass.
Under the silent correction model, you type the misspelled word, the correction is applied automatically in under 200 milliseconds, a brief visual indicator confirms the change, and you continue typing. There is no interruption. The error never persists on screen long enough to affect a message you might accidentally send before reviewing it.
Research on writing productivity consistently shows that interruptions to writing flow - even brief ones - reduce both the quality and quantity of output. A 2019 study published in the Journal of Writing Research found that writers who experienced more correction interruptions during drafting produced less coherent text than writers who worked uninterrupted. Silent correction preserves flow in a way that the annotation-and-review model cannot.
How does Charm's Spells feature work?
Charm is a native macOS menu bar application that provides real-time spelling correction through its Spells feature. The cyan glow that briefly appears around a corrected word signals that Spells has detected and replaced a misspelling. The correction happens in under 200 milliseconds - fast enough to be imperceptible during normal typing.
The mechanism that makes Charm work everywhere is CGEventTap - a macOS kernel-level API that allows an application to intercept and modify events from input devices before they reach any application. Because CGEventTap operates below the application layer, it is completely framework-agnostic. Whether the frontmost app uses AppKit, Chromium, Qt, or any other framework, Charm sees the same stream of keystrokes. The text framework used by the active app is irrelevant.
When you type a word and press a space or punctuation mark, Charm evaluates the completed word. If it is a misspelling, Charm injects the corrected form in its place using the Accessibility API. The whole process - detection, correction, injection - completes in under 200ms. The result is a writing experience where errors are eliminated before they can accumulate or be sent.
What Spells corrects
Spells handles standard spelling errors: transpositions ("recieve" to "receive"), missing letters ("defintely" to "definitely"), substitutions ("seperate" to "separate"), and common phonetic misspellings. It uses a personal dictionary that learns over time, so it does not over-correct technical terms, proper nouns, or vocabulary specific to your work.
Spells does not attempt to correct grammar or style - that is the role of Charm's Polish feature, which handles sentence-level grammar errors. The two features work independently and can be toggled separately from the Charm menu bar icon. This separation means you can run spelling correction alone if you prefer, without grammar suggestions changing your sentence structure.
Setting up spelling correction with Charm
Getting Charm running takes under two minutes. Here is the process from start to finish.
Step 1: Download Charm. Visit theodorehq.com/charm, click Get Charm, and complete the $9.99 one-time purchase. Download the .dmg installer from the link provided after purchase.
Step 2: Install the app. Open the .dmg file and drag Charm.app to your Applications folder. Eject the disk image when done.
Step 3: Launch Charm and grant Accessibility permission. Open Charm from your Applications folder. macOS will prompt you to grant Accessibility permission - this is required for Charm to read and correct text across apps. Click Open System Settings and enable Charm under Privacy and Security - Accessibility. Return to Charm.
Step 4: Enable Spells. Click the Charm icon in the menu bar. Toggle Spells on (it glows cyan when active). You can also enable Polish (grammar, blue) and Oracle (word prediction, purple) at this stage if desired.
Step 5: Test it. Open any app - your email client, a notes app, a messaging tool, a developer tool. Type a deliberate misspelling. Watch for the brief cyan glow as Charm replaces it with the correct form. Spelling correction is now active in every app on your Mac.
For more detail on each step, see the complete Charm setup guide.
Spelling correction use cases: who needs it most?
Spelling correction is useful for almost everyone who writes on a Mac, but some use cases benefit more than others.
Fast typists
Fast typists make more errors per unit time than slow typists, simply because speed and accuracy trade off against each other at high words-per-minute rates. A typist working at 80+ WPM will accumulate spelling errors faster than the red-squiggle model can prompt review. Silent real-time correction means errors are cleaned up as they happen, without any cognitive overhead. The typist never slows down.
Writers with dyslexia or spelling difficulties
For writers with dyslexia, spelling errors are frequent and the red-squiggle model is particularly disruptive - it draws constant visual attention to errors on shared screens and requires ongoing interruption of the writing process. Silent correction removes both the visual marker and the manual intervention. Writing continues without the repeated reminder of each error. Coverage in every app is especially important: dyslexic writers need assistance everywhere they write, not just in a browser or in one app.
Non-native English speakers
Writers working in English as a second or third language often have fluent grammar but inconsistent spelling - particularly around words with silent letters, doubled consonants, or vowel combinations that differ between languages. Silent correction handles these patterns automatically and predictably, reducing the cognitive load of monitoring spelling while composing.
Professionals communicating at volume
Professionals who send dozens of messages, emails, and documents per day face cumulative risk from unreviewed spelling errors. A single misspelling in a client email or a submitted document can undermine professional credibility disproportionately. A study by the Global Language Monitor found that grammar and spelling errors in professional communications reduce perceived competence ratings by an average of 15%. Silent correction eliminates this risk at the point of typing, before the message is sent.
How does Charm compare to macOS built-in spell check?
The table below summarises the key differences between the two approaches.
| Feature | macOS spell check | Charm Spells |
|---|---|---|
| Coverage | Native AppKit apps only | Every app on Mac |
| Correction model | Red squiggles, manual fix | Silent auto-correction |
| Speed | Manual (seconds per error) | Under 200ms, automatic |
| Grammar correction | No | Yes (Polish feature) |
| Word prediction | No | Yes (Oracle feature) |
| Privacy | On-device | On-device |
| Price | Free (included with macOS) | $9.99 one-time |
The macOS built-in option is adequate for users who write exclusively in native Apple apps and are comfortable with manual correction. For anyone who types in a broader range of apps - or who wants corrections to happen without interrupting their workflow - Charm's Spells feature is the more capable option.
For a full side-by-side breakdown, see our dedicated complete guide to autocorrect on Mac.
Spelling correction and writing privacy
All of Charm's correction, including Spells, happens entirely on your Mac. No text is transmitted to any server. No account is required. The corrections work offline. This matters for anyone who handles confidential material - medical records, legal documents, financial data, personal communications - where transmitting text to a cloud service is not acceptable.
macOS built-in spell check is also on-device. Cloud-based alternatives send your keystrokes to remote servers for processing, which introduces a privacy trade-off that many professional users cannot accept. Both Charm and the macOS built-in system avoid this exposure.
A 2024 survey by the International Association of Privacy Professionals found that 68% of enterprise employees were unaware that browser-based productivity tools transmitted keystroke data to third-party servers. On-device processing removes this risk entirely. For a detailed look at writing privacy on Mac, see our complete guide to writing privacy on Mac.
Frequently asked questions
How does spelling correction work on Mac?
macOS includes a built-in spell checker that underlines suspected misspellings in red inside native apps. It does not silently correct - you must right-click each underlined word and select the correct spelling manually. It has no coverage in apps that use non-native text rendering.
Why does Mac spell check not work in some apps?
macOS spell check is part of the AppKit text framework. Apps built on non-native frameworks (such as Chromium or Electron) bypass macOS text services entirely. Spell check has no mechanism to reach these apps. Charm solves this by operating at the kernel level via CGEventTap, below any app framework.
What is the difference between red squiggles and real-time spelling correction?
Red squiggles mark errors but require manual action: right-click, select the correct word, repeat for each error. Real-time correction (Charm Spells) detects and replaces the misspelled word automatically as you type, in under 200ms, with a brief cyan glow. No manual action is needed.
Does Charm work in apps where macOS spell check does not?
Yes. Charm uses the macOS Accessibility API and CGEventTap to intercept keystrokes at the kernel level, below any app framework. This means it corrects spelling in every Mac app - including productivity apps, developer tools, and messaging apps where macOS spell check cannot reach.
Is Charm's spelling correction private?
Yes. Charm processes all corrections on-device using local models. Your text never leaves your Mac. No internet connection is required for corrections to work, and no account or login is needed. This makes Charm safe for confidential or professional writing.