Groundbreaking Upgrade to SVG Is Approaching: Linked Parameters Are Almost Here

Published on:: November 4, 2026
Categories:: Wishlist, SVG, CSS⁹¹
Current music:: Tom Figgins —
Into the Maw
<a href="https://earthbornegames.bandcamp.com/track/into-the-maw-stereo-mix">Tom Figgins<span class="nowrap"> —</span> Into the Maw</a>
Current drink:: Yunnan tea

Introduction

“This feature will change everything” might be a bit of an exaggeration, but it, surely, will change a lot. In this post, I’ll report on what is happening right now with it, and why I think this is so exciting.

So! 🔜™️ the Linked Parameters spec’s implementation will land in all major browsers. Yes, in all of them!

What Are Linked Parameters?

Did you ever want to style an SVG file from outside? Like, when it is included in an <img>, or in a background-image, or in a border-image… Ok, the latter one is a stretch, but if you actually did try — join my club, haha.

But, really. If you are working on some design system, and you need to style some SVG, the chances are — you are using some overengineered library that embeds an SVG into your JSX, and then it is included in your React bundle or something, dynamically re-rendering whenever something needs to change in that SVG.

Megabytes of non-cached SVG code that has to be parsed and applied via JS. “Fun”!

What if there was a better way?

I won’t show any live examples in this post — I already provided a few in my article about one workaround for the absence of native linked parameters, and below I will link to many articles, explainers, and other links that could do this better than me.

In short, if you have some SVG, you will be able to mark some values there as coming from outside by the custom env () variables somewhere in SVG’s CSS:

rect {
	fill: env(--color, pink);
}

And then pass this variable from external CSS (there are other ways to do it; a declaration is my preferred way to do so):

.icon {
	background: url('my-icon.svg');
	link-parameters: param(--color: deepPink);
}

Note: an earlier versions of the spec were using param (--foo, bar), but in one of the recent CSSWG meetings we resolved to use the colon syntax, similar to how we now do this in if (), and how it will be in named arguments of custom functions and mixins — once they will eventually get them.

If you would try this in one of the experimental implementations, those could still use the older syntax.

The main three things I want to point at in this post are:

There are currently ongoing implementations for them in all major browsers.
There was a lot of interest from authors for them over time.
I have my opinion over what this can do for the web platform in general.

Ongoing Implementations

Firefox Nightly

It is already in Firefox Nightly! Here is the meta bug in bugzilla, opened on January 24th, 2023 by Sebastian Zartner.

The implementation started in March 2026 by Tyler Thrailkill, who picked up this after an open invitation from Emilio Cobos Álvarez to mentor someone working on this feature. How cool is that?

Blink

The work for Chromium was started by Divyansh Mangal and continued by Rahul Singh, following the corresponding Microsoft Edge Explainer that was published on June 19, 2026, and which, in turn, followed a Blink design plan.

Here is the The Chrome Status page for the feature.

The initial Chromium issue was opened on December 11, 2023 by Rune Lillesveen.

WebKit

There were also already many PRs merged from Taher Ali for WebKit, starting from July 28, 2026.

And — it was mentioned in the “Release Notes for Safari Technology Preview 254” on October 8, 2026! And it is listed as “supported” in WebKit CSS Feature Status, although I did not yet see it anywhere in the Safari Release Notes.

So, in a span of a few months, actual work started and is still actively ongoing in all major engines!

A Bit of History

This is something authors wanted for a long time!

The oldest proposal that I found was by Doug Schepers — a working draft from June 16, 2009!

Later, there was an unofficial proposal draft by Tab Atkins-Bittner, Daniel Holbert, and Jonathan Watt from September 14, 2018.

Another notable issue that brought it forwards was one opened by Lea Verou on January 28, 2024.

On March 24, 2025 I published my “Passing Data into SVG: Linked Parameters Workaround” article. The code inside is obnoxious, but I highly recommend reading the article itself — there are many fun bits inside. Not sure how much my article contributed to this being worked on right now, but it was at least mentioned in the Microsoft’s explainer, under “Evidence of developer demand”.

As usual, I mentioned many other people inside my article, and two notable mentions that I want to bring from it are those that led to me writing it article, from two very different points in time:

The spark in form of a Mastodon thread by Ivy Reese on January 31, 2025.
And the seed for my old draft in form of an article from June 3, 2023 by Estelle Weyl.

Another push was from a potential Interop topic issue opened by Ollie Williams on September 4, 2025. It was not accepted, which was kinda expected without any engines committing to it, but already mentioned above Divyansh Mangal submitted initial WPT tests on October 23, 2025.

Patrick Brosset mentioned linked parameters in the first point of his “What’s missing from SVG” article from June 22, 2026.

Dave Letorey contributed a link-parameters CSS property MDN reference in July 2026.

Vadim Makeev published “One SVG, nine posters: CSS linked parameters” article on September 23, 2026.

There was a “Linked parameters = arbitrary customizations!” section in the recent article by Lea Verou from October 11, 2026.

Finally, this year we got an Interop submission from Ollie as well, and while we don’t know if it will be accepted, given all browsers are working on this feature, I don’t think anyone will veto it.

It still might not get into Interop, as there were still plenty of spec questions, but hey, the chances are now much higher! And — I am pretty sure — if it will not be in the interop for the next year, then it will be in 2028.

So… 27 years after the initial proposal, we are finally getting it! And there is plenty of interest in them!

Why It Matters So Much

If it was not apparent, I am very excited about this thing. If you met me this year, there was a big chance I mentioned linked parameters to you once or twice. Why?

I think SVG is very underrated and underutilized. While CSS improved so much over the last few years, there are still things SVG can do better, or differently enough that it is still needed.

My big hope is that linked parameters will connect the success of the modern CSS with SVG, and unlock SVG as a tool to be used on a much bigger scale than it was used before. Not just for icons.

Being able to include SVG anywhere, without bundling, and make it fully styleable from outside… You could URL-encode it into your CSS (like with this awsome tool by Yoksel), and then adjust many things inside just by passing various CSS variables through.

Forget CSS functions and mixins (ok, don’t — they’re awesome), you could encapsulate some pretty nifty decorative logic inside such SVG file, and then share it in a fully configurable plain .svg file…

Of course, SVG has many problems (performance, anyone? I will always recommend you to learn from Ana Tudor from her work with SVG filters, like from her “SVG Filters Guide: Getting Started with the Basics” article, but the performance of them is pretty abysmal in all browsers), and I also hope that heightened interest to it could bring more attention from both authors and browser engineers, motivating everyone to pay much more attention to SVG and encourage finding ways to improve it for everyone.

This is not the last time you hear about linked parameters from me, so — keep an eye on my future posts and maybe even articles, and keep an eye on browser changelogs. They’re almost here, and it was never so exciting if you care about both CSS and SVG. Honestly.