Markdown Remove Images Keep Alt

Replace Markdown images with their alt text

Markdown Remove Images Keep Alt finds every inline Markdown image and swaps it for the description in its square brackets. ![Team photo at the 2025 offsite](photos/team.jpg "Offsite") becomes simply Team photo at the 2025 offsite. The image address and any title are dropped; the words that describe the picture stay where the picture was. On screen the tool is titled Markdown Remove Images.

Keeping the alt text rather than deleting the image outright matters. Well-written alt text often carries information the surrounding sentences rely on, for example "Chart showing sign-ups doubling in March". Throwing it away would leave holes in the meaning; keeping it turns a visual reference into readable prose.

When a text-only version is what you need

• Plain-text channels. Email newsletters sent as text, SMS, chat tools and ticket systems cannot show Markdown images and display the raw syntax instead.
• Content without hosted assets. When you move articles to a new platform before the images are migrated, broken image icons look worse than a short description.
• Accessibility review. Reading a document with images replaced by their alt text is a fast way to see what a screen reader user hears. Empty or vague descriptions stand out immediately.
• Language model and search input. Image URLs are long and meaningless to a text model. Alt text is short and descriptive.
• Word counts and print layouts. Removing image syntax gives cleaner text for counting, printing or pasting into a word processor.

Which image forms are recognised

The conversion pattern in the tool covers the common inline image forms:

Several images on one line are all converted. Ordinary links, written without the leading exclamation mark, are not touched.

Not converted

• Reference-style images such as ![Logo][logo] with a separate definition line.
• HTML images written as <img src="..." alt="...">.
• Images inside fenced code blocks. Anything between ``` or ~~~ fences is preserved, as the "How it works" note on the tool explains, so documentation that shows image syntax as an example stays correct.

Linked images and badges

README files often wrap an image in a link, for example a build badge: [![Build status](badge.svg)](https://ci.example.com). The tool replaces the inner image, so the line becomes [Build status](https://ci.example.com), an ordinary text link with the alt text as its label. That is usually exactly what you want: the badge becomes a readable link that still goes somewhere.

It all happens in your browser

The conversion is done with JavaScript on this page. Your Markdown is not uploaded and no server is involved, so unpublished articles and internal documents stay on your device.

Step-by-step

1. Paste your Markdown into the Input Markdown editor, or click Paste. Click Load sample to try a ready-made example.
2. Check the counters at the top: Input lines, Images found (outside code fences) and Code blocks skipped.
3. Click Remove Images in the header, or Remove under the editor.
4. Read the Clean Output editor. A toast confirms "Images replaced with alt text." or tells you no Markdown images were found.
5. Click Copy, or Download to save markdown-without-images.md.
6. Clear empties both editors.

Writing alt text that survives well

Because the output is only as good as your alt text, a quick pass over your images before converting pays off:

• Describe the purpose, not the file. "Screenshot of the Settings page with the Export button highlighted" beats "screenshot1".
• Fill in empty alt text where the image carries meaning. Empty alt text produces nothing in the output, which is right for decorative spacers but wrong for a chart.
• Avoid "image of" and "picture of". In the converted text these phrases read awkwardly.
• Keep it short. One clear sentence is easier to read inline than a paragraph; long explanations belong in the body text or a caption.
• Check punctuation. An image placed mid-sentence becomes words mid-sentence. If the alt text ends with a full stop, you may get doubled punctuation.

Example: turning a product update into a text email

Suppose a release note reads:

We redesigned the dashboard. ![New dashboard with the usage chart on the left](img/dash.png "Dashboard") The usage chart now loads twice as fast.

After conversion it becomes:

We redesigned the dashboard. New dashboard with the usage chart on the left The usage chart now loads twice as fast.

The description reads as part of the paragraph, but it lacks a full stop. That is typical: alt text is written as a caption, not as a sentence. A quick edit, adding a full stop or turning it into "(Image: New dashboard with the usage chart on the left.)", makes the text email read naturally. Images placed on their own line, as most images in documentation are, need no such edit, since the alt text simply becomes a short line of its own.

The Images found counter is also useful here as a checklist. If it reports twelve images and your email will be read by people who cannot see any of them, those twelve alt texts are now the only description they get, so it is worth reading each one once.

What not to do

• Do not expect the image URLs to be kept anywhere. If you need a list of them, copy them out first, or use Markdown Links to URLs, which turns images into their addresses instead.
• Do not use it to clean HTML images. Run Markdown Remove HTML for tags, keeping in mind that it removes the alt text too.
• Do not rely on it for reference-style images; convert those to inline images first.

Limitations

• Inline images only; reference-style and HTML images are not changed.
• The alt text is matched up to the first closing square bracket, so alt text that itself contains ] is not recognised.
• An image whose syntax wraps across two lines is not matched, because the tool works line by line.
• An unclosed code fence protects the rest of the document, so images after it are not converted.
• No size limit is set in code; very large files depend on your browser.

Frequently Asked Questions