Snipping Tool Text Actions vs a Clipboard Shelf

Snipping Tool Text actions read pixels and return characters. An asset shelf keeps images (and sometimes files) ready to drop. Treating OCR as a visual library is how error dialogs get copied as messy paragraphs while the original bitmap is already gone.

Microsoft documents Text actions as an on-device Optical Character Recognition feature inside a captured snip. It is a good feature. It is not history.

What Text actions actually do

After a snip is open in Snipping Tool:

1. Select Text actions.
2. Windows runs OCR locally. Microsoft’s Snipping Tool article states that all text recognition processes are performed on the device.
3. Select specific words, or use Copy all text.
4. Use Quick redact when the snip contains email addresses or phone numbers that should not leave the image as selectable text.

The result of Copy all text is text on the clipboard. The result of a normal snip is a bitmap. Those are different formats. Windows clipboard history can store both, as text and as bitmap, but they occupy separate slots if both are copied.

Text actions do not:

• Rename the PNG
• Pin the image
• Build a stack of related shots
• Survive as an image after the next copy unless the bitmap was saved or still sits in history
• Replace a password manager or a notes app

For a longer OCR walkthrough that includes PowerToys, see How to Copy Text Out of an Image on Windows.

What an asset shelf is for

A shelf, tray history, or even a well-kept Screenshots folder answers a different question: *where is the picture.*

Microsoft’s history rules still apply to anything copied after Text actions: 25 items, 4 MB each, text / HTML / bitmap, unpinned list cleared on restart. A copied novel from a long article can be text small enough to stay in history. A 4K snip of the same article can miss history entirely.

When Text actions are the right tool

Use Text actions when the pixels are a vehicle for words:

• An error code in a dialog that cannot be selected
• A serial number printed in a screenshot from a vendor
• A table in a locked PDF page (expect imperfect columns)
• A slide in a video call where copy is blocked

Then paste the text into the ticket, not the bitmap, unless visual context is required. If both are required, save the PNG first, then run Text actions, then copy the text. Order matters. Saving first means the image still exists after the text copy overwrites the live clipboard.

Pin neither the secret nor the full dump. Pins survive restart. A one-time code that sits pinned in Win+V is a gift to the next person at the desk. See How to Pin Items in Windows Clipboard History.

When the image is the deliverable

Use the file when:

• Engineering needs to see the exact control that failed
• Design is comparing two builds
• The OCR would mangle layout (icons, mixed CJK and Latin, dense dashboards)
• The next step is Figma, Photoshop, or a slide

Path:

1. Win+Shift+S or Snipping Tool New.
2. Confirm auto-save or press Ctrl+S. See Win+Shift+S Not Saving What You Think It Saves.
3. Annotate in Snipping Tool if a single arrow is enough. Otherwise mark up in the destination app so the original PNG stays clean.
4. Place or attach the file. Do not rely on Win+V as the archive.

ShareX’s after-capture list includes Recognize text (OCR) as a separate task from Save image to file and Copy image to clipboard. That is the same split, with more automation. Official task names live on getsharex.com.

History is not a shelf

Win+V shows recent text, HTML, and bitmaps that fit the cap. It does not show files. It does not filter “images only” in a useful designer way. It does not search last month. After a reboot, unpinned cards are gone.

That is enough for “paste the hex copied ten minutes ago.” It is not an asset shelf.

A real shelf, in plain language, would:

• Keep named images on disk or as durable local cards
• Let a mouse drag the file into Explorer, Slack, or a canvas
• Survive restart without a special pin ritual for every shot
• Stay out of cloud sync unless that is an explicit product

Windows does not ship that. Optional local tools exist. Edge-Drop is a hover shelf for staging copies and dragging them out. It is not OCR, not Snipping Tool, and not required if Pictures\Screenshots plus good filenames already work.

Accuracy and language

OCR is statistical. Expect errors on:

• Thin UI type against busy backgrounds
• Single-character codes (O versus 0)
• Mixed scripts
• Text that is part of a screenshot of a screenshot

Microsoft states recognition runs locally. That is a privacy plus. It is not a promise of perfect tables. For a structured table, prefer selecting the real text or exporting CSV from the source app. A snip plus Text actions is the fallback when the UI forbids selection.

If the extracted text will be pasted into a terminal or a code editor, scan for curly quotes and stray line breaks. History will store whatever OCR emitted, mistakes included.

Worked example: an error dialog

1. Reproduce the error.
2. Win+Shift+S around the dialog. Confirm auto-save or Ctrl+S so error-checkout.png exists.
3. Open the toast. Run Text actions > Copy all text.
4. Paste the text into the ticket description.
5. Attach error-checkout.png to the ticket.
6. Delete the OCR text from Win+V if it contains an email or order ID that should not linger.

The ticket now has words *and* pixels. History holds neither as the long-term store.

What “Copy all text” does to the live clipboard

It replaces the live clipboard with text. The bitmap is no longer Ctrl+V. If auto-save was off and the toast is dismissed, the only image may already be gone. That is the strongest argument for saving before OCR.

If both must stay available in history, copy the image first (already true after the overlay), turn history on so the bitmap is listed, *then* run Text actions so the text becomes a second card. Both cards still count toward 25 and still vanish at restart if unpinned.

A clean split for a workday

1. Capture the UI.
2. If the picture matters, confirm a PNG exists.
3. If the words matter, run Text actions and copy text *after* the file exists.
4. Paste text into the note. Place the file into the design or ticket.
5. Clear or avoid pinning secrets.

Click to Do on Copilot+ PCs can also copy recognized text or images from the screen. That is another action layer, still not a library. Microsoft documents Copy for both text and images in Click to Do. Treat it like Text actions: extract, then file the original if the pixels matter.

If you are choosing on product lines rather than the built-in panel, Screenshot Folder vs Clipboard History for Designers is the honest split.

Related reading

• How to Keep the Last Ten Screenshots Handy
• Screenshot Folder vs Clipboard History for Designers
• How to Get a Screenshot Into Photoshop Without Recopying
• How to Paste Without Formatting on Windows

Sources

• Use Snipping Tool to capture screenshots (Microsoft Support) — Text actions, Copy all text, Quick redact, on-device OCR, Ctrl+C.
• Using the clipboard (Microsoft Support) — 25 items, 4 MB, text/HTML/bitmap, restart wipe.
• Click to Do: do more with what's on your screen (Microsoft Support) — Copy recognized text or images; not a history product.
• ShareX — after-capture OCR listed separately from save and clipboard copy.
• How to use clipboard history in Windows 11 (Microsoft) — enabling history and pins.

Copy. Stack. Drop.

Transform your clipboard into an interactive edge shelf. Stack, pin, and drag assets into any app with zero friction.

Download for Windows Get from Microsoft Store

How to Install Guide · First 10 Minutes Guide · Drag & Drop Guide · Edge-Drop vs Win+V · Support