ScreenMD — 4–8× fewer tokens than a screenshot

Screenshots cost 4–8×
more tokens than Markdown.

ScreenMD grabs any page as clean, structured text — one hotkey, straight to your clipboard. No file dialogs. No Downloads folder hunting. Just paste into Claude, ChatGPT, or Gemini and go.

Install Extension Download for Mac Windows — Coming Soon
Press M for Markdown, J for image, L for region, ; for region OCR — pick any mode, one keystroke.

Modern AI chats are token-hungry. Screenshots are bulky, OCR-dependent, and lose structure. Copy-pasting strips formatting and drags in navigation junk. ScreenMD gives the model exactly what it needs: the content, in the structure the page already has, with nothing extra.

How it works

Three seconds. Zero friction.

ScreenMD runs quietly in your menu bar and browser. Press a hotkey — structured Markdown hits your clipboard.

STEP 01
⌘⌃M

Press the hotkey

From any app or browser. ScreenMD detects what's focused automatically — no mode switching.

STEP 02
extracting…

Content is extracted

Full page DOM for browsers — not just the visible viewport. Accessibility tree for native apps like Word, PDF, and Mail.

STEP 03
✓ copied

Markdown hits your clipboard

A toast confirms it's ready. No files, no upload dialogs, no hunting for a screenshot in Downloads.

STEP 04
⌘V

Paste into any AI tool

Claude, ChatGPT, Gemini, Cursor — rich, structured context at a fraction of the token cost.

Five capture modes

Pick the shape of context
your AI actually needs.

Different tasks call for different captures. ScreenMD gives you six purpose-built modes — three work in both the browser extension and the Mac app; three are Mac-app exclusive (region image, region OCR, and PDF/Word page range).

⌘⌃M
Browser + Mac

Markdown

Clean structured text — headings, tables, code, links — to your clipboard. 4–8× fewer tokens than a screenshot. The default for sending pages and documents to ChatGPT, Claude, or Gemini.

⌘⌃J
Browser + Mac

Image

Full-page screenshot to your clipboard. Scroll-and-stitch in Chrome handles long pages without duplicating sticky headers. Ideal when layout, charts, or visual design is the point.

⌘⌃K
Browser + Mac

Both

Markdown and image, captured together. Markdown lands on your clipboard for paste; the screenshot saves to Downloads. Use it when you want structured text plus a visual reference.

⌘⌃L
Mac app only

Region

Drag a crosshair to capture exactly the area you care about. The cropped image goes straight to your clipboard — no Preview round-trip, no editing.

⌘⌃;
Mac app only

Region OCR

Drag a crosshair over any image, chart, or scanned content — Vision OCR extracts the text and sends it straight to your clipboard as Markdown. Same crosshair as region, text output instead of image.

⌘⌃P
Mac app only

Page range

For PDFs and Word docs, a one-line dialog asks which pages to capture — pre-filled with your current page. Type 6-8 or all. Captures exactly what you need, nothing more.

Example output — Markdown mode

This is what the AI actually receives.

Here's what ⌘⌃M produces: structured headings, code blocks, and links — exactly as they exist on the page. No OCR guesswork, no sidebar noise.

Without ScreenMD ~2,000 tokens
screenshot.png · 1.4 MB

Drag file → upload dialog → OCR errors → AI confused by sidebar → expensive

With ScreenMD ~350 tokens
## Browser Context
- Tab: CORS errors in Node.js
- URL: stackoverflow.com/...

### Top Answer
Add cors middleware...

```app.use(cors())```

Press ⌘⌃M⌘V in AI chat → done. Structured, clean, cheap.

clipboard contents — markdown
Markdown copied
## Browser Context
- **Browser:** Chrome
- **Tab Title:** How to fix CORS errors in Node.js
- **URL:** stackoverflow.com/questions/...
- **Captured:** 04/24/2026 10:32 AM

## Page Structure
### Question
How do I resolve a CORS policy error when calling my Express API
from a React frontend?

### Top Answer (482 votes)
Add the cors middleware to your Express app...

## Code Samples
```
const cors = require('cors');
app.use(cors());
```

## Links
- [cors npm package](https://npmjs.com/package/cors)

What it does

The same hotkey.
Every source covered.

Two pieces, one workflow. The browser extension handles any webpage. The Mac app handles Word, PDF, Mail, and any focused window.

Browser Extension

Clean Markdown extraction

Headings, paragraphs, tables, links, and code blocks — nav chrome, footers, modals, and tracking params filtered out automatically.

Browser Extension

Full-page screenshots

Scroll-and-stitch handles sticky headers without duplicating them. Works on any site — no allowlist, no per-site config required.

Browser Extension

Reddit-aware capture

Captures posts and top comments via Reddit's public API for a clean, complete thread view — no screenshot gymnastics.

Mac App

Word & Pages support

Extracts document text directly — not a screenshot of it. Heading structure, body copy, and formatting all preserved.

Windows equivalents coming Q2 2026.

Mac App

PDF page-range picker

Preview and Skim auto-detect your current page, or pop a dialog to pick a range. Captures exactly what you need, nothing more.

Mac App

Apple Mail capture

Pulls message body, sender, subject, and date as clean Markdown. Feed whole email threads to your AI without copy-paste.

Both

Consistent hotkeys

⌘⌃M, J, and K work identically in the extension and the Mac app — no mode switching. ⌘⌃L (region), ⌘⌃; (region OCR), and ⌘⌃P (page range) are Mac-app exclusive.

Both

Output exactly where you want it

Clipboard for ready-to-paste context. Downloads folder for screenshots you want to keep. Your call, per shortcut.

Both

No setup after install

Grant permissions once. ScreenMD stays out of your way until you need it — no accounts, no configuration.

Hotkey cheat sheet

One hand. Six shortcuts.

M, J, and K work in both the browser extension and the Mac app. L, ;, and P are Mac-app exclusive — they need native screen access and document APIs the browser can't reach. Hotkeys auto-adapt to your OS: Windows and Linux users see Alt+Shift in place of ⌘⌃.

Shortcut Action Browser Mac app
M
Markdown to clipboard
J
Screenshot to clipboard
K
Markdown to clipboard + full-page screenshot to Downloads
L
Region screenshot to clipboard — drag crosshair to select
;
Region OCR to clipboard — drag crosshair, get text instead of image
P
PDF / Word page-range picker — selected pages to Markdown on clipboard

How people use it

Situations you'll recognize.

Built because the gap between what's on screen and what's in the AI chat was costing too many tokens and too many steps.

You're mid-conversation in Claude

You need that Stack Overflow answer. ⌘⌃M — the page hits your clipboard as clean Markdown. Paste. Done. No tab hunting, no file picker, no 1.4 MB screenshot the AI has to OCR.

You have a 40-page PDF

But you only need pages 12–18. The page-range picker pre-fills your current page. Type the range, clipboard gets exactly those pages — nothing more, nothing less.

You just found the article

A screenshot would cost ~2,000 tokens and hand the AI a wall of sidebars and ads. Markdown costs ~300 tokens and gives it just the content. Same result, 6× cheaper.

Privacy

Your captures never leave your machine.

Your captured content stays local — clipboard or Downloads folder, nowhere else. The only thing our server ever sees is your license key during activation.

  • Captures are written only to your local clipboard or Downloads folder
  • No telemetry, no crash reporting, no analytics pipeline
  • No tracking of what you capture, when, or how often
  • Reddit captures fetch public Reddit JSON directly from your browser
  • Paid licenses contact our server only to activate, refresh, or deactivate a device — never with capture content
Read the full privacy policy →

Pricing

Free to try. One price, forever.

No subscription. No per-device fees. One license unlocks the browser extension, the Mac app, and the Windows app (when it ships) on every device you own.

Free

Try the browser extension with no signup.

$0
  • Browser extension only
  • 10 captures per week
  • Every browser capture mode unlocked
  • No account required
  • Mac & Windows apps require a license

Lifetime license

Unlimited captures across every platform you use.

$9.99 one-time
  • Unlimited captures, forever
  • Browser extension, Mac app, and Windows app (included free when it ships)
  • Up to 5 active devices
  • All future updates included
Get the lifetime license

Get started

Up and running
in under two minutes.

Install one or both — they work independently. You don't need the Mac app to use the extension, and vice versa.

Browser Extension

Handles any webpage on any platform. One install covers Chrome, Edge, Brave, and Arc.

  1. 1 Click Install Extension below — opens the Chrome Web Store, which works in every Chromium browser
  2. 2 Pin the ScreenMD icon to your toolbar (optional but handy)
  3. 3 Press ⌘⌃M on any page — that's it

Works with

Chrome Edge Brave Arc

Edge users: when prompted to allow extensions from other stores, click Allow — one-time per browser profile.

Install Extension

Free on the Chrome Web Store · works in every Chromium browser

Mac Desktop App

Handles Word, Pages, Preview, PDF, Apple Mail, and any focused window. Lives quietly in your menu bar.

  1. 1 Download ScreenMD.dmg below
  2. 2 Drag to Applications and open it
  3. 3 Grant Accessibility + Screen Recording permissions once
  4. 4 Press ⌘⌃M anywhere — done

Requires

macOS 12 Monterey+
Download for Mac

Windows Desktop App Coming soon

Native capture for Word, Excel, PowerPoint, Outlook, Acrobat, and Edge PDF — same hotkeys, same clipboard-first workflow. Targeted for Q2 2026.

  1. 1 Use the browser extension on Windows today — every browser capture mode works
  2. 2 Sign up for early access on the native app — we'll email when the signed installer is ready

Will support

Windows 10+ Windows 11
Get early access

FAQ

Common questions.

Do I need both the extension and the Mac app?

No. The extension handles browser pages on its own. The Mac app handles native documents on its own. Install one, install both — your call.

Does ScreenMD upload my browsing data?

No. Captured content never leaves your machine — it goes straight to your clipboard or Downloads folder. The only network calls ScreenMD ever makes are fetching public Reddit JSON for Reddit captures, and (for paid users) contacting our license server to activate, refresh, or deactivate a device. Capture content is never sent anywhere.

What about paywalled or logged-in pages?

Whatever you can see in your browser, ScreenMD can capture. No special setup required — it reads what's already rendered for you.

Will this work with my app?

The Mac app natively supports Word, Pages, Preview, Skim, and Apple Mail. Any other focused window falls back to a clean window screenshot automatically.

Can I change the hotkeys?

The browser extension's hotkeys are configurable in chrome://extensions/shortcuts. The Mac app's hotkeys are fixed in this version.

Is it free?

The browser extension has a free tier — 10 captures per week, no account, no signup. The Mac and Windows apps require a license. A $9.99 one-time purchase unlocks unlimited captures across the extension, the Mac app, and the Windows app on up to 5 devices, forever. No subscription.

What happens when I hit the free weekly limit?

The browser extension shows a one-time notice with an upgrade link. The counter resets every Sunday. Every browser capture mode works on the free tier — the license is purely a usage cap (and the unlock for the Mac and Windows apps).

If I buy a license, does it cover both the browser and Mac?

Yes. One license unlocks both. Paste the same key into the extension and the Mac app — each counts as one of your 5 device slots.

One hotkey between your screen and your AI.

Free tier: 10 captures a week, no signup. $9.99 one-time unlocks unlimited use across browsers, Mac, and Windows. No subscription, ever.

Install Extension Download for Mac