CS2 Video Settings: What Pros Actually Use
The CS2 video settings pro players actually run - resolution, scaling, max FPS, texture, shadow, and Reflex - computed live from cs2pedia's dataset of 1,000+ pro profiles, with honest denominators on every setting.
Most CS2 video-settings guides hand you the same copy-paste list - “set textures to Low, cap your FPS, done” - with nothing behind it. This one is different in one specific way: every number below is computed at build time from cs2pedia’s dataset of 1,000+ pro profiles, and where a setting is only recorded for a fraction of the field, the figure says so on its face.
That honesty matters, because some of these settings are well documented across the pro field and others are recorded for barely a third of it. Treating a sparse field as a full-field consensus is the most common mistake in this category, and the captions below exist so you never have to take a percentage on faith.
A clear majority of the pro field plays on a 4:3 aspect ratio. That single fact frames almost everything else on this page, so we start there - then walk the rest of the settings in roughly the order you meet them in the in-game Video menu.
Resolution and Aspect Ratio
Aspect ratio is the shape of the rendered image - 4:3 is the squarer, CS:GO-era format; 16:9 is your monitor’s native widescreen. Among the pros who recorded it, 4:3 is the clear majority, with 16:9 a distant second and 16:10 and 5:4 splitting most of the remainder.
The single most common pro resolution is 1280x960 - the canonical 4:3 resolution - used by well over half the field that recorded a resolution, with 1024x768 and the native widescreen 1920x1080 trailing.
The short version of why: a 4:3 image stretched across a 16:9 panel makes enemy player models wider, and the lower pixel count frees up frames. That is the whole argument in one sentence, and it is genuinely contested - there are real reasons to stay on native 16:9, and the field is not unanimous.
The full decision, including the case for each side and exactly how to make the switch, lives in its own guide: CS2 4:3 vs 16:9. We won’t re-litigate the verdict here.
What we’ll add is the qualifier the copy-paste lists skip: the data above is a snapshot of pro configs, not a live per-event feed, and the share is computed only against profiles that actually recorded an aspect ratio - see the caption denominators. The aspect-ratio field is well populated, so this one is close to full-field consensus. Several settings further down are not.
Scaling Mode
If you run a non-native resolution like 1280x960 on a widescreen monitor, scaling mode decides what fills the leftover screen space.
- Stretched distorts the 4:3 image to fill the whole panel (wider models, no black bars).
- Black Bars keeps the image at true 4:3 geometry with unused bars on the sides.
- Native is for players already on their monitor’s native ratio.
Among pros who recorded a scaling mode, Stretched is the runaway plurality, with Native and Black Bars splitting the rest roughly evenly.
No Scaling is folded into Native.
Stretched is what produces the wider-model effect; Black Bars gives you the lower pixel count and FPS headroom of 4:3 without the horizontal distortion. The choice between them is the core of the 4:3 vs 16:9 guide, so we’ll leave the trade there. The data point worth keeping here is simply that the field overwhelmingly prefers Stretched among those who reported it.
Display Mode
Display mode has three options:
- Fullscreen (exclusive fullscreen, where CS2 owns the whole display).
- Fullscreen Windowed (a borderless window the size of your screen).
- Windowed.
Among the pros who recorded it, this is the closest thing to unanimity on the entire page - the field runs Fullscreen almost without exception.
The reason is input latency. Exclusive Fullscreen lets the game bypass the desktop compositor and present frames directly, which shaves a small but real amount of latency versus a borderless window.
Fullscreen Windowed is more convenient if you alt-tab constantly, and on some systems the latency gap has narrowed - but the pro field has not moved off exclusive Fullscreen, and the recorded share makes that clear. Set it to Fullscreen unless you have a specific multi-monitor workflow reason not to.
Max FPS
The fps_max console variable caps your frame rate. Set it to 0 and the game renders as many frames as your hardware allows (uncapped); set it to a number and frames are capped there.
Here is where disclosure stops being a formality and starts being the whole point: max_fps is the most sparsely recorded setting on this page - it is blank for the large majority of profiles. So read the figure below strictly as “among the minority of pros who reported a max_fps value,” not as a statement about the field as a whole.
Uncapped = fps_max 0; Capped = any positive value.
Among that recorded minority, uncapped (fps_max 0) is the most common choice. The logic is that more frames means lower average frame latency and a fresher image at the moment you click, so players with CPU headroom simply let it run.
The competing view caps fps_max slightly above the monitor’s refresh rate to keep frame pacing even and the CPU cool - a reasonable choice, but a minority one, and the specific capped values are too thinly spread to call a single “correct” cap. If you have the headroom, uncapped is the safe default; if your frametimes are erratic, a cap a little above your refresh rate is a defensible alternative.
fps_max 0Texture and Shadow Quality
These are the two settings people most expect to be a simple “set it to Low” - and the data tells a more interesting story, which is exactly why disclosing the denominators matters.
Model / Texture Detail controls the resolution of in-game textures. Among the roughly two-thirds of pros who recorded it, Low is the clear majority. The reasoning is sound: lower textures cost nothing in visibility (player models are still fully visible) and free up VRAM and frames. This is the closest thing to a true “competitive default” on the page.
Global Shadow Quality is where the conventional “everything Low” advice breaks down. Shadow is recorded for well under two-thirds of the field, so treat the split as the view of the pros who reported it - but among those, the plurality actually runs High, not Low.
That surprises people, so here is the rationale: shadows are information. An enemy’s shadow can reveal a player you can’t yet see directly - around a corner, off a wall, behind boxes on specific map angles. Higher shadow quality renders those shadows more completely, so a meaningful slice of the field treats shadows as a visibility tool rather than an FPS cost and runs them up.
There is no single right answer: Low buys you frames and a cleaner image; High buys you shadow-based information. The data shows the field genuinely split, with High leading among those who reported it - which is the opposite of the lazy “all Low” recommendation, and a good example of why the denominator-honest approach is worth the trouble.
NVIDIA Reflex Low Latency
NVIDIA Reflex reduces the render queue - the backlog of frames the CPU has prepared for the GPU - which trims the latency between your input and the frame that reflects it. It has three states:
- Disabled
- Enabled
- Enabled + Boost (Boost also keeps the GPU clocked up to shave a little more latency when you are GPU-bound).
This field is recorded for under half the dataset, so the split below describes only the pros who reported a Reflex value - it is not a full-field reading.
Among that recorded group, Enabled and Disabled are close to even, with Enabled + Boost the smaller third. The practical read: Enabling Reflex has essentially no downside on a supported NVIDIA card, so it is reasonable to turn it on.
Boost specifically only helps when you are GPU-bound - when the GPU, not the CPU, is the bottleneck. CS2 at competitive settings is usually CPU-bound on a decent graphics card, which is why Boost stays a minority choice: most pros simply aren’t GPU-limited often enough for it to matter.
That points at a budget reality: if you’re GPU-bound on an older card, the lever that actually moves your latency and frame rate is the hardware, not a single in-game toggle. A monitor and GPU that can keep up will help far more than chasing Reflex Boost - see the best monitor guide for the displays the pro field actually runs.
Shader, Ambient Occlusion, and FSR
The remaining graphics options are quick calls, and they all point the same direction - toward the lowest setting that doesn’t cost you information. Each of these fields is recorded for only part of the dataset, so read each share against its own caption denominator rather than as a full-field number.
Shader Detail affects surface and lighting effects. Among the pros who recorded it, Low is the strong majority - it costs frames for visual flourish that does nothing for your aim.
Ambient Occlusion adds soft contact shadowing where surfaces meet. It is sparsely recorded, but among those who reported it, Disabled is the plurality - it’s a visual nicety with a frame cost and no competitive upside.
FSR (FidelityFX Super Resolution) is AMD’s upscaling tech - it renders at a lower internal resolution and upscales, trading sharpness for frames. Among the pros who recorded it, it is almost universally left at its highest-quality / disabled state; the field is already running low enough native resolutions that upscaling buys little.
The pattern across all three is the same one that runs through this whole page: turn off anything that adds visual polish without adding information.
Digital Vibrance
Digital vibrance is the one setting on this list that isn’t a CS2 video option at all, which is why it doesn’t appear in the dataset - it lives in your GPU’s control panel, not the game. It increases color saturation, which makes enemy player models pop against muted map backgrounds (Dust2’s tan, Mirage’s sandstone).
A large share of the competitive community runs it elevated, but because it’s a driver setting rather than a per-profile game config, we don’t have a dataset distribution for it and won’t invent one.
- On NVIDIA: open the NVIDIA Control Panel → Display → Adjust desktop color settings → set Digital Vibrance higher (many players land somewhere in the 70–90% range, but this is taste, not a measured pro figure).
- On AMD: the equivalent lives under AMD Software → Display → Custom Color → Saturation.
Push it until enemies stand out without the image looking cartoonish - there’s no “correct” value, and unlike the in-game settings above, we have no pro distribution to anchor it to, so treat any specific number you see online (including ours) as a starting point, not a benchmark.
A Note on the Data
Every distribution on this page is computed at build time from cs2pedia’s dataset of 1,000+ active pro profiles - the same database that powers the rest of the site - not a hand-typed list that goes stale the moment a player tweaks a setting.
The captions tell you exactly how many profiles recorded each field, because that is the honest way to present it: aspect ratio, resolution, scaling, and display mode are recorded for the large majority of the field and read close to consensus.
Max FPS, shadow, Reflex, ambient occlusion, and FSR are recorded for a smaller slice, and their shares describe only the pros who reported them. A sparse field dressed up as full-field agreement is the exact thing this page refuses to do.
The usual caveat applies: configs change between events and the dataset is a snapshot, not a live per-match feed, so treat the exact shares as current-but-not-instant. For the per-command technical detail, Valve’s own settings documentation is the primary source.
If you want to dig further: see the aggregate settings stats for the whole dataset at a glance, or browse individual player profiles to see exactly what any given pro runs. For the aspect-ratio decision in full, read CS2 4:3 vs 16:9; to make sure your display can keep up with the frames these settings unlock, see the best monitor guide.