Every platform that ships a new display capability ends up answering the same question twice, once for the people who own the hardware and once for the people who have to write software against it, and on Quest 3 the second answer is the more interesting one. Meta's developer documentation on display refresh rates now describes a headset that can run its panel far past 120 Hz, and it lays out the ceiling, the conditions attached to it, and the frame budget that comes with each step up. It also carries a line about not requiring any of it.
Meta Documented a 240Hz Quest 3, Then Told Developers Not to Require It

The number that matters to somebody who already owns the headset is 207. Meta's page states that on Quest 3 the display hardware "accepts any integer refresh rate from 72 Hz through 207 Hz," and that an app can ask for any of them through the same APIs it already uses to request 90 or 120. The word integer is doing real work there. It means the panel is not restricted to a short list of blessed modes the way earlier Quest hardware was, so a developer targeting, say, 144 or 160 is choosing a number rather than picking from a menu.
Where the 240 comes from
Above 207 Hz the rules change, and they change in a way that anyone writing a spec sheet should understand before quoting the bigger figure. Reaching 240 Hz requires developer mode and two commands that switch on display scaling, and display scaling is exactly what the name suggests: the frame is rendered at a lower resolution, handed over as a smaller payload, and upscaled by the display hardware on its way to the lens. Meta does not soften this. "Display scaling trades image quality for bandwidth," the documentation says, and the resulting image "is not identical to a natively rendered full-resolution image, and the difference is visible in fine detail such as small text and thin geometry."
Which makes the honest summary something like this. Quest 3 has a native ceiling of 207 Hz at full resolution, reachable through ordinary APIs, and a 240 Hz mode that costs you resolution and lives behind a developer toggle. Both are real. They are not the same product decision, and a store page that advertises the second without the first would be selling something the headset does not quite do.
The frame budget is the part worth taping to a monitor
Meta publishes the arithmetic alongside the rates, and the arithmetic is where enthusiasm meets a mobile SoC. At 72 Hz an app has 13.9 ms to produce a frame. At 90 Hz that drops to 11.1 ms, and at 120 Hz to 8.3 ms. At 207 Hz the budget is 4.8 ms. At 240 Hz it is 4.2 ms.
If you have shipped anything on a standalone headset you already know what 8.3 ms feels like to hold on a bad frame with passthrough running and a user who keeps wandering across the room. Four point two is not that problem with a harder number attached. It is a different problem. Meta's own guidance draws the line for you: extended rates suit apps that "render a light scene and benefit from low latency and smooth motion, such as rhythm games, competitive shooters, sports and racing titles, and simple high-contrast environments," while "an app with a dense scene, heavy post-processing, or a high draw call count is a poor candidate."
That is a narrower target than the headline implies, and it is worth reading next to the work that has gone into making 72 and 90 hold steady on this same silicon. The industry spent years building frame extrapolation to stop headsets dropping below their floor. The ceiling moving to 240 does not retire any of that.
Quest 3S is not invited

Extended refresh rates are, in Meta's words, "exclusive to Meta Quest 3." Quest 3S tops out at 120 Hz because it uses a different panel. Quest 2 also stops at 120, Quest Pro at 90, and the original Quest never went past 72. So the split runs straight through a lineup Meta currently sells side by side, and the cheaper headset that shares a generation name with the Quest 3 does not share this.
For a developer that is the awkward part. Extended rates are not a platform capability you can assume across Horizon OS. They are a property of one specific SKU, which means any code path built around them needs a fallback that is not an afterthought, and any marketing built around them needs an asterisk naming the model.
The sentence to read before planning around it
Meta says it directly on the same page: "Do not assume that a future headset will support rates above 120 Hz, and do not make an extended rate a requirement for your app to run."
Platform holders rarely hedge that hard on their own new capability, and the hedge is the most useful thing in the document. It is the difference between a feature you can adopt and a feature you can depend on. Anyone who shipped through the Android fragmentation years has met this shape before, where a flagship device gains something genuinely good, the docs quietly note that it may not carry forward, and the studios that treated it as a baseline spent the next two years writing compatibility code. Meta is telling developers in advance which category this belongs to, which is more than the industry usually gets.

What has to be true on the headset
The documentation gives one hard requirement: "Extended refresh rates require HorizonOS v2.7 or later." Road to VR reported on August 31 that the higher rates had begun appearing on retail Quest 3 units, naming Horizon OS 2.7 as the build carrying them, with 207 Hz available normally and 240 Hz reachable after enabling developer mode.
Set against the rest of 2026 the timing is a little funny. Valve spent this year moving in the opposite direction, cutting its Steam Frame Verified requirement from 90 fps down to 72 so that more of the catalog could clear the bar. Meta has raised a ceiling almost nobody will reach. Those are not contradictory positions. One is about what a certification badge should promise, the other is about what a panel is physically capable of, and the gap between them is roughly the whole practical story of VR performance right now.
Alex Reeves covered the first 240Hz AR gaming glasses back in May, where the refresh rate was the product and buyers paid for it directly. Here the same number is a documented property of hardware people already own, which makes it a better deal and a smaller event at once.
For anyone shipping in six months, the practical read is short. Treat 207 as the real ceiling, treat 240 as a demo, target it only if your scene is genuinely light, and write the fallback first. And if the app is going to advertise a rate at all, say which headset it means.
