Bypass Medium Paywall - Working late 2023 - Greasy Fork, Violentmonkey, Tampermonkey - Click the RAW button to install

medium.user.js

// ==UserScript==
// @name        Medium Paywall Bypass
// @namespace   Violentmonkey Scripts
// @run-at      document-start
// @match       *://*.medium.com/*
// @match       *://medium.com/*
// @match       *://*/*
// @grant       none
// @version     2.3
// @inject-into content
// @updateURL   https://gist.githubusercontent.com/mathix420/e0604ab0e916622972372711d2829555/raw/medium.user.js
// @downloadURL https://gist.githubusercontent.com/mathix420/e0604ab0e916622972372711d2829555/raw/medium.user.js
// @website     https://freedium.cfd
// @author      Mathix420, ZhymabekRoman
// @description Don't forget to remove `@match` filters you don't want.
// ==/UserScript==

// initCall is telling us if we need to inject the title observer
function mediumRedirecter(initCall = false) {
  // I miss typescript...
  const mediumPostUrlProperty = ((document.head || {}).querySelector ? document.head.querySelector('meta[property="al:android:url"]') : {}) || {}

  if (
      // Allow seeing original articles that were already redirected to freedium.
      !window.location.href.endsWith('#bypass') &&
      // Do not redirect when editing on medium.
      !window.location.href.includes("/edit?source=") &&
      // Detect if we are on a medium website (regardless of the domain)
      (mediumPostUrlProperty.content && mediumPostUrlProperty.content.includes('medium://p/'))
    ) {
    window.location.href = 'https://freedium.cfd/' + window.location.href;
  } else if (initCall && /(.*\.|^)medium\.com$/.test(window.location.host)) {
    new MutationObserver(function(mutations) {
      if (mutations[0].target.textContent) mediumRedirecter();
    }).observe(
        document.querySelector('title'),
        { subtree: true, characterData: true, childList: true }
    );
  }
}

mediumRedirecter(true);

Thank you very much for this Gist, it works like a charm!

You're welcome!

Adding // @inject-into content is required for Violentmonkey.

Brilliant, works like a charm! Thank you 🚀

Thanks!!!

Thank you for your work. If you also write on medium, currently editing your publication results in an error page. One might choose to apply the change.

20c19
<   if (window.location.href.endsWith('#bypass')) {
---
>   if (window.location.href.endsWith('#bypass') || window.location.href.includes("/edit?source=")) {

To prevent the redirect that causes you to reach the error page.

Nice addition @rvente, thanks!

I think the @run-at property should be set to document-body, because with document-start it sometimes does not perform the redirects. Problem occurs in Firefox, everything is fine in Chrome.

@ZhymabekRoman interesting, and yes it might be possible that the body is not fully loaded when document.head is run I guess.
But I can't find any reference to document-body online, where did you find this value? Maybe it is defaulting to document-end if it's recognized as a bad value. Or most likely I just didn't find a great source of documentation.

But I can't find any reference to document-body online, where did you find this value? Maybe it is defaulting to document-end if it's recognized as a bad value. Or most likely I just didn't find a great source of documentation.

Hmmm, yeah, I think document-body is only Tampermonkey specific : https://www.tampermonkey.net/documentation.php?locale=en#meta:run_at

Violentmonkey doesn't support it: https://violentmonkey.github.io/api/metadata-block/#run-at

to automatically redirect to the freedium site:

// ==UserScript==
// @name        Medium Paywall Bypass
// @namespace   Violentmonkey Scripts
// @run-at      document-start
// @match       *://*.medium.com/*
// @match       *://medium.com/*
// @match       *://*/*
// @grant       none
// @version     1.5
// @inject-into content
// @updateURL   https://gist.githubusercontent.com/mathix420/e0604ab0e916622972372711d2829555/raw/medium.user.js
// @downloadURL https://gist.githubusercontent.com/mathix420/e0604ab0e916622972372711d2829555/raw/medium.user.js
// @website     https://freedium.cfd
// @author      Mathix420, ZhymabekRoman
// @description Don't forget to remove `@match` filters you don't want.
// ==/UserScript==

(setInterval(function () {
  // 1. Allow seeing original articles that were already redirected to freedium.
  // 2. Do not redirect when editing on medium.
  if (window.location.href.endsWith('#bypass') || window.location.href.includes("/edit?source=")) {
    return;
  }

  const mediumPostUrlProperty = document.head.querySelector('meta[property="al:android:url"]')
  if ((mediumPostUrlProperty || {}).content && mediumPostUrlProperty.content.includes('medium://p/')
      && window.location.href.match(/^https?:\/\/(www\.)?(medium\.com\/|[\w-]+\.medium\.com\/|[\w-]+\.[\w-]+\/).*/
)) {
    window.location.href = 'https://freedium.cfd/' + window.location.href;
  }
}), 1000);

Hi @ZhymabekRoman

Violentmonkey doesn't support it

Surprisingly I just saw it in the settings of my Violentmonkey. Anyway I'm surprised that using document-body as the docs says The script will be injected **if** the body element exists. maybe it's a typo but all the other descriptions says when. It's pretty strange.
Sorry for being this picky on this change, but as it drastically increase the time of redirect I'm a bit reluctant to it. Also it works perfectly on my end with violentmonkey, would you consider trying with this one ? Might also be better for your privacy.

---

Hello @lonelam, thanks for your recommendation, but I don't understand your change can you give me some context? As the condition checking for medium meta tag meta[property="al:android:url"] already allows us to redirect every medium article (even those not hosted by the medium domain).

I'm using Violentmonkey, maybe the other extensions are breaking headers loading (I have about ~45 extensions).

Hi @ZhymabekRoman

Violentmonkey doesn't support it

Surprisingly I just saw it in the settings of my Violentmonkey. Anyway I'm surprised that using document-body as the docs says The script will be injected **if** the body element exists. maybe it's a typo but all the other descriptions says when. It's pretty strange. Sorry for being this picky on this change, but as it drastically increase the time of redirect I'm a bit reluctant to it. Also it works perfectly on my end with violentmonkey, would you consider trying with this one ? Might also be better for your privacy.

Hello @lonelam, thanks for your recommendation, but I don't understand your change can you give me some context? As the condition checking for medium meta tag meta[property="al:android:url"] already allows us to redirect every medium article (even those not hosted by the medium domain).

The Medium site is an SPA and navigating to a new page will not trigger a 'load' event. And therefore, the script will not triggered if I navigate from a medium's page to another,
So a manual refresh is needed before I jump to freedium, do the detection intervally will prevent the extra manual operation.

@ZhymabekRoman okay, let me know if I still need to change the code !

---

@lonelam Oh I see! Sorry I missed the setTimeout part from your code! I also noticed this behavior earlier, I was thinking about implementing something like this https://stackoverflow.com/q/6390341/9799292

@ZhymabekRoman okay, let me know if I still need to change the code !

@lonelam Oh I see! Sorry I missed the setTimeout part from your code! I also noticed this behavior earlier, I was thinking about implementing something like this https://stackoverflow.com/q/6390341/9799292

That's an awesome performance impovement, Listening the events like locationchange or hashchange will be more economically and more effective for the detection.

@lonelam I just updated the script, it can now detect title changes (only when using medium.com domain as custom domains does not seems to host SPAs) and trigger a redirect if needed!
PS: small bug fix, make sure to be on V2.1

Achtung!

Sooo, it was going to take a while, but now we have it. Our whole Github organization is not public for now.

Reddit community, that was beginning all of that also gone - https://www.reddit.com/r/paywall/comments/15jsr6z/bypass_mediumcom_paywall/

We have moved to Codeberg - https://codeberg.org/Freedium-cfd

Medium, thank you >.

Thanks all Freedium-cfd members!

@mathix420 Can you fix this error? Sometimes this error appears:

TypeError: can't access property "querySelector", document.head is null

@ZhymabekRoman I've fixed the type error, but if the head tag is missing it won't be able to detect the medium page before js load.
Sorry for the delay, I got lot of work these days!

@mathix420 Thank you! I really appreciate your efforts

I think you can simplify the below lines using Optional chaining (?.):

First, completely remove

- const mediumPostUrlProperty = ((document.head || {}).querySelector ? document.head.querySelector('meta[property="al:android:url"]') : {}) || {}

Then inside the first if statement, replace

- (mediumPostUrlProperty.content && mediumPostUrlProperty.content.includes('medium://p/'))
+ document.head?.querySelector('meta[property="al:android:url"]')?.content?.includes('medium://p/')

Personally, I'm not going to use // @match *://*/*. So I put this into a detectMediumWebsite() function and commented it out.

Thanks for your work.

P.S. By the way, are you sure that the MutationObserver is actually registered?

Can you remove? // @match *://*/*

do not make sense to be for all sites.

do not make sense to be for all sites.

Medium has too many sub-domains

Medium has too many sub-domains

// @match *://*.medium.com/* Should be enough for matching sub-domains.

Should be enough for matching sub-domains.

What about devopsquare.com, blog.devops.dev, blog.stackademic.com, ai.plainenglish.io, bettermarketing.pub and etc? It's impossible for us to say how many Medium sites there are.

You are right. I didn't know those type of sites exist. But they are just different domains, technically speaking.

To recap:
// @match *://*.medium.com/* Is enough for matching sub-domains.
// @match *://*/* Is needed for matching any possible domain who happens to be a Medium website.

freedium is offline so unfortunately this script doesn't work anymore for now

All works as expected. Yeah, we made some codebase refactor, to improve speed and some big bug fixes. And there were some minor downtimes.