On November 12, 2026, the W3C published the first Working Draft of CSS Color HDR Module Level 1, the spec that teaches the cascade about screens brighter than your living-room lamp. It's an early draft, not a shipping feature, but it's the first time the CSS Working Group has written down, formally, what a web page is supposed to do when the display underneath it can hit 1,000 or 10,000 nits instead of the 100-nit ceiling every CSS color function has quietly assumed since rgb(). Two days earlier, Firefox 160 had already shipped a small, flagged piece of the same puzzle: an experimental colorSpace option on the Canvas 2D API that lets a canvas render in display-p3 instead of plain sRGB. Teams shipping production interfaces on tight timelines rarely have a spare engineer to track a working draft like this one, which is exactly the kind of gap you close when you extend your dev team with people who already read W3C minutes for a living.
CSS Color HDR Module Level 1 extends CSS Color 4 and 5 rather than replacing them. It adds three HDR color spaces tied to the Rec. 2100 broadcast standard: rec2100-pq (Perceptual Quantizer, up to 10,000 cd/m² of peak luminance), rec2100-hlg (Hybrid Log-Gamma, built to adapt to whatever brightness the viewer's screen happens to be running), and rec2100-linear for linear-light math. Alongside those it defines device-independent spaces built for perceptual uniformity at high brightness, ICtCp, Jzazbz and JzCzhz, reachable through new functions like ictcp() and jzazbz().
The more immediately useful part, for anyone who isn't a color scientist, is the control surface: dynamic-range-limit lets a stylesheet say whether an element is allowed to use the display's full peak brightness (no-limit), stay capped to ordinary SDR levels (standard), or sit somewhere in between (constrained), with dynamic-range-limit-mix() for blending those states across a transition. The spec also pins down vocabulary the industry has used loosely for years: HDR Reference White is fixed at 203 cd/m², and "headroom" is defined as the logarithmic measure of how far a color can go above that white point before clipping.
The Firefox 160 release notes, published November 10, 2026, list a disabled-by-default gfx.canvas.colorspace.enabled flag that adds a colorSpace option to HTMLCanvasElement.getContext() and OffscreenCanvas.getContext(), supporting "srgb" and "display-p3", plus a matching colorSpace property on ImageData. That's a narrower, already-partially-supported color space (Safari and Chrome have shipped display-p3 canvas rendering for a while), not the new HDR spaces from the W3C draft. But it's the same underlying problem: browsers rendering color correctly requires both a wide-gamut pipeline and a way for authors to opt into it deliberately, and the CSS spec is now writing the rulebook that canvas, image formats and eventually the rest of CSS color will all have to agree on.
Nothing in this draft is ready to ship in production CSS today: it's a Working Draft, the earliest formal stage, and the spec text itself says it may be "updated, replaced, or obsoleted" at any point. For most teams the practical takeaway is smaller and more immediate: if you're already serving wide-gamut photography, video, or canvas-rendered graphics with display-p3, start treating color space as an explicit decision in your rendering pipeline rather than an implicit sRGB assumption. The vocabulary in this draft, reference white, headroom, dynamic range limit, is also a useful checklist when you're reviewing a design system that claims HDR support but hasn't actually defined what happens when a non-HDR screen renders the same page.
Watch for prefers-color-scheme-style media feature queries tied to display capability; the draft's authors have signaled that detecting HDR support at the CSS level is a near-term follow-up, separate from this first module.
Working Draft is the first of several W3C maturity stages: Candidate Recommendation and Proposed Recommendation still have to happen before this is a finished standard, and CSS Working Group drafts routinely take one to three years to move through that pipeline. Treat late 2027 at the earliest as a realistic window for browser implementations to start landing behind flags, mirroring how Firefox's own display-p3 canvas support is still experimental years after Safari and Chrome shipped the equivalent. Track the spec's GitHub repository rather than this article for the authoritative status; working drafts can and do stall.
If your product already touches high-fidelity color (product photography, medical imaging, design tooling, anything rendered to canvas or WebGL), this is worth a line item on next year's roadmap, not this sprint's. The honest move right now is to make sure whoever owns your rendering pipeline is aware the spec exists, so the first production decision isn't made under deadline pressure once a browser ships it for real.