How to convert text to Markdown from any web page

If you search for "convert text to markdown", you will find a lot of tools that solve part of the problem.

Some convert pasted HTML. Some fetch a public URL and return Markdown. Some save the current page from your browser. They all sound similar until you try them on the pages you actually use: logged-in docs, private Notion pages, member-only newsletters, LMS pages, support portals, internal dashboards, or long articles you want to send to ChatGPT without dragging along menus and footer links.

I tested Web2MD on that exact job: take the readable text from a web page and turn it into clean Markdown I can paste into ChatGPT, Claude, or Cursor. The short version: if the page is public, you have options. If the page only exists inside your browser session, a browser-side converter is often the practical answer.

Web2MD is a Chrome extension for converting web pages to Markdown. It runs locally in your browser, includes a token counter, and has one-click send-to-AI actions. The free tier gives you 3 conversions per day. Pro is $9 per month. The honest limits are also simple: it is Chrome-only right now, and the free tier is not unlimited.

What "convert text to markdown" usually means

Markdown is plain text with light formatting. Headings use #, links use brackets and parentheses, lists use hyphens, and code uses backticks.

That matters because AI tools handle Markdown well. A clean Markdown copy of a page usually gives a model the article structure, headings, links, tables, and code without dumping every navigation item into the prompt.

Here is a small example of raw page content after conversion:

# Refund policy

You can request a refund within 30 days of purchase.

## How to request a refund

Send your order number to support@example.com and include the reason for the request.

## Exceptions

- Gift cards are not refundable.
- Enterprise contracts follow the terms in the signed agreement.

That is much easier to read than copied browser text with cookie banners, sidebar labels, and random spacing. It is also easier for an AI tool to summarize or answer questions about.

The problem with public URL converters

Server-side readers are useful. Jina Reader, for example, is fast and convenient when the page is public. Firecrawl is strong when you need crawling, scraping workflows, or structured extraction through an API. These tools are good at what they are built for.

The catch is access.

A server-side tool can only fetch what its server can reach. If the content requires your browser login, a session cookie, a paid account, or a page state created after you click around, the server may see a login screen instead of the article.

That is where browser-side conversion is different. Web2MD runs where the page is already open: in your browser. If you can read the page in Chrome, the extension can work from that page. In my testing, this was the main practical advantage over URL readers.

That does not mean Web2MD replaces every scraping tool. If you need to crawl 5,000 public pages on a schedule, use an API-first product like Firecrawl. If you need a quick Markdown view of a public URL, Jina Reader is still handy. If you want to convert the page you are looking at, especially a logged-in page, Web2MD fits better.

What I tested

I tested Web2MD on a mix of pages I actually use for AI research and writing:

• A public blog post with headings and code
• A logged-in documentation page
• A long newsletter opened in the browser
• A support article with nested lists
• A page with lots of navigation links around the main text

The output was not perfect on every page. No converter is. Complex tables still need a quick skim. Pages that render content in unusual shadow DOM setups can be messy. Some sites fill the page with repeated UI text, and any converter has to guess what is content and what is chrome.

But for ordinary articles, docs, and help pages, the Markdown was clean enough to paste directly into an AI chat. I usually checked the headings, removed one or two irrelevant lines, then sent it.

Why clean Markdown helps AI tools

Copy and paste from a browser often gives you text, but not useful structure. A model can still read it, but the prompt becomes harder to scan and easier to misread.

Markdown preserves structure without adding much noise.

For example, a documentation section might come out like this:

# Install the CLI

Install the package with npm:

```bash
npm install -g example-cli

Authenticate

Run the login command:

example login

The browser opens and asks you to approve access.

Verify the install

example --version

That structure is useful in ChatGPT, Claude, and Cursor because the model can see what is a heading, what is a command, and what is explanatory text. The token counter in Web2MD is also useful here. Before sending a long page to an AI tool, I can see roughly how large the converted Markdown is and decide whether to trim it.

## Web2MD compared with other options

MarkDownload is probably the closest browser-extension comparison. It is open source and has been around for a while. It is a good choice if you want a simple clipper-style extension and like configuring templates.

Jina Reader is great for public web pages. I use tools like it when I need a fast Markdown version of a URL and do not need browser login state.

Firecrawl is stronger for developer workflows: crawling, extraction, APIs, and automation. It is more than a one-page converter.

Web2MD is different because it focuses on AI workflows inside the browser:

- It works on pages your browser can access, including logged-in or paywalled pages you are allowed to view.
- It runs locally, so you are not sending the source page to a third-party reader just to convert it.
- It has a built-in token counter before you paste into an AI tool.
- It supports one-click send-to-AI for tools like ChatGPT, Claude, and Cursor.
- It has a free tier with no API key setup.

That privacy point is worth being precise about. Browser-side conversion reduces the need to send the page to a conversion server. If you then send the Markdown to ChatGPT or Claude, you are still sharing that Markdown with the AI provider you choose. Web2MD does not magically make external AI tools private. It just keeps the conversion step local.

## When I would use Web2MD

I would use Web2MD when the source page is already open in Chrome and I want to use it with an AI tool.

Typical cases:

- Summarizing a long article in ChatGPT
- Sending product docs to Claude for troubleshooting
- Giving Cursor clean documentation context
- Saving a paywalled article as Markdown for personal notes
- Turning a logged-in help center page into a prompt
- Checking token count before pasting a long source

I would not use it as my first choice for bulk crawling. I also would not use it if you need Firefox or Safari support today. Web2MD is a Chrome extension, and that matters if your workflow is built around another browser.

## A simple workflow

My usual flow is:

1. Open the page in Chrome.
2. Click Web2MD.
3. Review the Markdown preview.
4. Check the token count.
5. Copy it or send it to the AI tool.
6. Ask the model a specific question instead of just saying "summarize this".

That last step matters. Clean Markdown helps, but your prompt still matters. A better prompt is usually specific:

"Read this Markdown page and extract the setup steps. Keep commands exactly as written. Call out anything that looks version-specific."

That works better than asking for a generic summary.

## A note on limits and pricing

Web2MD has a free tier of 3 conversions per day. That is enough to test it or use it occasionally. If you convert pages all day, you will hit the limit. Pro is $9 per month.

I like that the limit is clear. I do not like surprise credit systems where a page costs a mystery number of credits because it was longer than expected. Here, the tradeoff is easy to understand: light use is free, regular use is paid.

The bigger limit is browser support. If you need non-Chrome browsers, this is not the right tool yet.

## Bottom line

If you only need to convert a public URL to Markdown, Jina Reader, Firecrawl, and other web readers may be enough. If you want an open-source browser clipper, MarkDownload is worth trying.

But if your real task is "convert the page I am looking at into clean Markdown for an AI tool", Web2MD has a practical edge. It runs in the browser, works with pages behind login when you have access, keeps conversion local, shows token count, and sends the result into the AI workflow without an API key.

You can try the Chrome extension from the Web2MD site: [Web2MD](/). Start with the free 3 conversions per day and see whether the Markdown is clean enough for the pages you actually use.