Google brought Android Dev Summit back this year after a four-year absence — it runs October 28 and 29 at Google's Mountain View headquarters, with an extended edition in London — and the published agenda did something Google almost never does: it named Android 18 outright, before any developer preview exists. The line that got picked up across 9to5Google, Android Authority, PhoneArena and half a dozen other outlets in the last week of September was a promise of "powerful new graphics features," and a single Jetpack Compose session on the agenda gives away what two of them are: mesh gradients and progressive blurs. If you build Android apps or design for them, the Android 18 graphics features teased in that agenda are worth reading now rather than after the keynote — because one of them is already shipping in Compose today, and the other is the kind of thing that changes how you'd scope a design system six months out.
What the Dev Summit agenda actually reveals
The session title is unusually specific: "Elevate your apps with the latest graphics features in Jetpack Compose and Android 18." Its description names mesh gradients, progressive blurs, and other new graphics capabilities — and, worth noticing, pairs them directly with performance, battery life, and frame rates. Google is introducing a set of visual effects and the cost of running them in the same breath. That pairing is usually a tell that the effects are expensive enough to need the caveat attached up front.
The rest of the summit splits across four tracks — AI experiences, Android XR, Apps foundation, and Tools and performance — with an Android leadership keynote that's the likely venue for anything broader about the release. What Google has not said is when Android 18 actually arrives: there's no confirmed developer preview date and no codename yet, and this agenda line is the first official acknowledgment the version exists at all. Treat any specific date circulating right now as speculation and plan against Google's recent release cadence instead.
Mesh gradients and progressive blur, concretely
A conventional gradient interpolates between stops along a single axis — linear from top to bottom, or radial out from a point. A mesh gradient blends colors across a grid of control points instead, so color can move in two directions at once and bend around the canvas. Practically, it's the difference between the flat two-tone background you can build today and the soft, multi-directional color washes that have dominated app and web design for the past couple of years. Teams currently fake these by exporting a PNG from a design tool and shipping it as a drawable, which costs download size, can't respond to theme or state changes, and looks wrong the moment it's stretched across an aspect ratio nobody tested.
Progressive blur varies blur radius across an area rather than applying one uniform amount. The canonical use is a scrim: content stays sharp at the top of a surface and blurs out gradually toward the bottom, so a title or a navigation bar stays readable over a photograph without a hard-edged dark overlay slapped across it. Done uniformly, blur reads as a modal state — something is behind this, and you can't touch it. Done progressively, it reads as depth, which is a different and much more useful design primitive.
Progressive blur already shipped in Compose — most teams missed it
Here's the part that makes this more than summit-watching: progressive blur is not something to wait for Android 18 to get. Compose's blur modifier already accepts a radius spec, so BlurRadiusSpec.verticalGradient(startRadius = 0.dp, endRadius = 24.dp) gives you a surface that's sharp at the top and 24dp-blurred at the bottom today, in one line, with no shader code. Chris Banes — who wrote Haze, the library most Compose teams reached for to get this effect — published a post this year titled, pointedly, "Compose has progressive blur. Do you still need Haze?" That question answers itself for a lot of codebases.
If you need effects past what the built-in modifier covers, the layer underneath is still available: RenderEffect, plus AGSL for custom fragment shaders, wired into Compose by creating a runtime shader effect and converting it with asComposeRenderEffect(). A healthy set of libraries sits on that foundation — Haze, skydoves' Cloudy (which exposes progressive modes like CloudyProgressive.TopToBottom() and runs on Kotlin Multiplatform), PrismalAGSL for liquid-glass treatments with graceful degradation, and imla for real-time hardware-accelerated blur. The gap Android 18 looks set to close is mesh gradients, which have no first-party equivalent right now and are genuinely awkward to approximate.
The API floor is the real design constraint, not the API
The constraint that decides whether any of this reaches your users isn't whether the Android 18 graphics features are good. It's minSdk. RenderEffect — the mechanism underneath hardware-accelerated blur on Android — landed in API 31. Anything below that needs a fallback path, and a platform-level mesh gradient introduced in Android 18 will sit even further up the version ladder, which means years before it's a safe baseline rather than a progressive enhancement. Any design that treats blur or mesh color as load-bearing needs a defined answer for devices that can't render it: a solid scrim, a flat gradient, a different composition entirely.
This lands hardest on exactly the projects where it's least convenient. Institutional EdTech deployments — the kind we build under D.TEN — frequently run on low-end, school-issued hardware several Android versions behind, where both the API floor and the frame budget are real. GPU blur is not free; applying it to a scrolling surface on a budget device is one of the more reliable ways to turn a smooth list into a janky one. Google pairing these graphics features with frame rates and battery life in the session description is the same warning from the other direction.
The honest sequencing, then: treat progressive blur as something you can adopt now with a fallback, and treat mesh gradients as a design direction to prototype rather than a dependency to plan a release around.
Planning a graphics effect for mid-range hardware
Visual effects have a cost, and the people most affected are those on modest phones. A short routine keeps you honest.
- List the screens that would use the effect, and which of them scroll or animate.
- Define the fallback first. What does the screen look like without blur or a mesh gradient? It should still be readable and attractive.
- Test on a real mid-range device and watch frame times while scrolling.
- Limit the blurred area. A small blurred surface is far cheaper than a full-screen effect.
- Offer a reduced-effects path when the device or the user's settings suggest it.
| Question | Why it matters |
|---|---|
| What is our minSdk? | It decides which devices can use platform effects |
| What is the fallback below API 31? | Hardware-accelerated blur needs a newer API level |
| Does the effect sit on a scrolling surface? | Blur on a list is the most common source of jank |
| What is the contrast of text over the effect? | A pretty scrim still has to be readable |
What we'd actually recommend
Don't wait for the October 28 keynote to start. If you have a surface where text sits over imagery — a media detail screen, a hero header, a bottom navigation bar over scrolling content — replace the hard scrim with Compose's built-in progressive blur and see it on a real mid-range device, not just an emulator. If you're already carrying Haze or a similar library purely for this one effect, check whether the platform modifier now covers your case and drop the dependency if it does.
Before the summit, write down your project's minSdk and what each blur-dependent screen degrades to below API 31. That single exercise is what converts "we should use the new Android 18 graphics features" from an aesthetic conversation into a shippable plan. For anything on the mesh gradient side, prototype the look in your design tool now so you know whether it's worth the wait — but don't commit a release to an API that has no announced developer preview, let alone a stable version.
And when the agenda lands, the session to actually watch is the Compose graphics one rather than the keynote highlights reel. The keynote will tell you what Android 18 can do; the Compose session is where the version floors, the performance costs, and the fallback story get spelled out — and those are the three things that determine whether any of it ships in your app.
Frequently asked questions
Sources

Written by
Hamza Fazal
Founder, DEESU
Muhammad Hamza Fazal is the Founder of DEESU. An Android and full-stack web developer and digital marketer based in Islamabad, he founded DEESU in May 2022 and leads its engineering operations, product strategy, and client software delivery.
More from HamzaFree tools for this topic
Need help with this?