Best FiveM Settings for FPS and Graphics

For a balanced FiveM setup, start with native resolution, High textures, Normal shadows and grass, MSAA off, and modest distance scaling. Use Normal textures instead when graphics memory is tight. Then test the settings on the server you actually play—not just in an empty street.

The aim is to keep the game responsive when you drive through a busy intersection, enter a custom interior or approach a crowd. Client scripts can affect FPS independently of graphics quality, which is why lowering every setting does not necessarily solve the problem. (Cfx Documentation)

This guide gives you a starting preset, explains what to change first and shows you how to check whether a change helped.

Version note: the presets below use GTA V Legacy’s graphics settings. They are not a tested Enhanced preset. For build-specific differences, consult Cfx.re’s FiveM for GTAV Enhanced documentation.

Best FiveM graphics settings: a balanced starting point

Before changing anything, take screenshots of your current settings. That gives you a reliable way back if the new configuration looks worse without improving performance.

Use the following as a suggested test preset, not a guarantee of a particular frame rate.

Setting	Recommended starting point
Screen Type	Fullscreen; compare borderless separately
Resolution	Your monitor’s native resolution
Refresh Rate	Your monitor’s normal, supported refresh rate
VSync	Off while benchmarking
FXAA	On; turn off if you dislike the softness
MSAA	Off
NVIDIA TXAA, where available	Off
Texture Quality	High, or Normal when VRAM is limited
Shader Quality	High
Shadow Quality	Normal
Reflection Quality	Normal
Reflection MSAA	Off
Water Quality	High
Particles Quality	Normal
Grass Quality	Normal
Post FX	Normal or High
Motion Blur Strength	Zero, according to preference
Anisotropic Filtering	8× or 16×
Ambient Occlusion	Off initially
Tessellation	Off initially
Population Density	Low to medium
Population Variety	Low to medium
Distance Scaling	Low to medium
Advanced Graphics	Extra effects off; extended-distance sliders at minimum
Frame Scaling Mode	Off for the initial native-resolution comparison

A few details matter here. Normal is the lowest selectable level for several Legacy settings, including textures, shadows and grass; there is no separate “Low” option for those. Population Variety affects the variety of loaded models and VRAM use, rather than simply reducing the number of people in the world. Anisotropic filtering improves angled textures at relatively little cost in NVIDIA’s original GTA V testing. (NVIDIA)

For visual comparisons, NVIDIA’s GTA V graphics guide remains useful for understanding Legacy settings. Its hardware benchmarks are historical, however—not predictions of performance on a modern FiveM server.

Find the bottleneck before lowering everything

Graphics settings help most when they reduce the work that is actually holding up the next frame. A CPU bottleneck and a GPU bottleneck need different responses: the GPU can be waiting for work from the CPU, or the CPU can be waiting for the GPU to finish rendering. Intel’s explanation of gaming bottlenecks covers this distinction. (Intel)

When the GPU is the likely limit

Temporarily lower the resolution while keeping everything else unchanged. Make sure VSync and any frame limiter are not holding the result at a fixed ceiling.

If FPS improves substantially in the same scene, that is evidence that rendering load matters. If it barely changes, resolution is probably not the most useful setting to sacrifice next. This is a diagnostic clue, not a complete diagnosis. (Intel)

For the next comparison, try reducing MSAA, grass, shadows or reflections, one setting at a time. These are better candidates than indiscriminately turning down every option. (NVIDIA)

When the client’s CPU or scripts are the likely limit

Compare a quiet street with a busy location, then repeat an action such as opening the inventory, entering an interior or approaching a gathering.

A repeatable drop tied to a particular action is worth reporting. Cfx.re’s profiler can identify client script execution associated with FPS problems; lowering resolution cannot remove that script work. (Cfx Documentation)

Try lower population density and distance scaling, but do not expect population sliders to remove connected players or repair an expensive resource.

When graphics memory is the concern

Try a lower texture-quality setting before sacrificing resolution. Texture quality and memory capacity are related, but lower textures are not automatically a large FPS improvement when enough VRAM is available. (NVIDIA)

Treat missing textures as a symptom to investigate—not proof that your graphics card is inadequate. Record whether the problem follows you across servers or appears at one particular location.

When the problem is connection lag

Use the performance overlay to check ping and packet loss as well as FPS. These are separate measurements: a high frame rate does not tell you whether your connection is healthy. (Cfx Documentation)

If the camera remains smooth but interactions arrive late, investigate the connection or server response before lowering image quality.

How to benchmark FiveM without misleading yourself

A useful comparison needs repeatable conditions. Testing one preset at an empty airport and another outside a crowded hospital tells you very little about the settings.

Turn on the FPS counter

Press F8 and enter:

cl_drawfps true

For a broader overlay, use:

cl_drawperf true

The second command shows FPS, ping, packet loss, CPU usage, GPU usage and GPU temperature. Both are documented user commands; you do not need developer mode to use them. (Cfx Documentation)

You can hide the overlays with:

cl_drawfps false
cl_drawperf false

See our FiveM FPS counter guide for the command reference and troubleshooting.

Repeat a short, representative route

1. Record the conditions. Note the server, approximate player count, resolution and settings. Keep your usual background applications consistent.
2. Let the initial loading settle. Then drive a route that includes a busy street and a location where you commonly notice problems.
3. Run the route more than once. Keep the vehicle, camera view and weather as similar as the server allows.
4. Change one setting. Restart when the game requires it, then repeat the same comparison.
5. Keep changes that produce repeatable improvements. Revert settings that damage image quality without a clear benefit.

A small difference on one pass is a reason to test again, not to declare a winner.

Pay attention to pauses, not just the highest number

At a steady 60 FPS, each frame takes approximately 16.7 milliseconds. At 120 FPS, it takes approximately 8.3 milliseconds. Those figures describe a consistent pace; an occasional long frame can interrupt it.

The basic overlay is useful for quick checks, but it is not a complete frame-time recording. For deeper diagnosis, the Cfx.re profiler can help developers investigate spikes associated with client scripts. (Cfx Documentation)

For your own comparison, note whether the same pause happens when you turn a corner, enter a building or open a menu. “The inventory causes a hitch every time it opens” is more useful than “sometimes I get low FPS.”

Settings for lower-end PCs

Start with the balanced preset, then make it more conservative: Normal textures and shaders, Normal shadows and grass, MSAA off, and low population and distance scaling. Keep advanced graphics disabled during testing.

Try native resolution first. If the resolution test shows that the GPU is limiting performance, compare a lower resolution with the same aspect ratio. Do not force a lower-resolution image when it produces no meaningful improvement.

Once the result is stable, test High textures again. Keep them only if the image improves without bringing back the stutter you were trying to eliminate.

Choose your target from the route you actually play. Start by testing whether the system can maintain 60 FPS through ordinary gameplay rather than chasing a much higher number in an empty area. That is a target for evaluation, not a promise that every server will sustain it.

Settings for better graphics on a stronger PC

Start from the same baseline rather than selecting the maximum value everywhere.

My suggested upgrade order is textures and shaders first, then shadows and reflections. Test grass separately on a countryside route, and leave MSAA and extended-distance options until you have measured the other changes.

Use two comparisons: a demanding location that represents normal play, and a scene where you can judge the visual improvement. Keep the setting only when the improvement is worth the performance trade-off to you.

Afterward, test VSync and your monitor’s synchronization setup under normal gameplay conditions. “Off while benchmarking” is a testing choice, not a rule that you must leave it off permanently.

Fix FiveM stutter, texture loss and slow loading separately

These problems can look similar, but they need different tests. Do not change graphics settings, drivers, cache files and background applications all at once.

Check storage when loading is the problem

Where possible, keep GTA V and FiveM’s working data on an SSD. Faster storage can shorten loading, while storage bottlenecks can contribute to pauses as data is loaded. It will not make a slow graphics card render faster. (Intel)

Also identify which loading stage is slow. Waiting in a queue, downloading resources and waiting for character initialization are different problems. Our guide to reducing FiveM loading times explains how to separate them.

Test optional visual effects and overlays

Run one comparison without optional graphics modifications, recording overlays and applications you do not need for the session. Keep everything else unchanged.

If the result improves, re-enable them individually. That gives you an identifiable cause instead of leaving you with a long list of disabled features and no explanation.

The same principle applies to GPU drivers. Record the version before updating, and avoid combining a driver change with a complete settings overhaul.

Do not clear the cache as routine maintenance

Keep cache clearing for a specific troubleshooting reason. Do not empty it between normal benchmark runs: that changes the loading conditions and can create additional downloads. (RPScripts)

A useful report includes the exact error or visual problem, where it occurs and whether reconnecting changes it. “I clear the cache every week” is not a performance measurement.

Report location-specific texture problems

When roads, buildings or interiors disappear, repeat the route and note the location, vehicle, direction of travel and current texture setting.

Ask whether other players can reproduce it. If several people see the same problem at the same place, pass that evidence to the server team before making further global changes to your PC.

Server owners should inspect the affected assets, including their textures, levels of detail and file integrity, rather than assuming every missing surface is a player settings problem. See our FiveM server optimization guide for that workflow.

Server owners: separate resource cost from graphics cost

A server resource may include code that runs on the player’s machine. That client-side execution is different from server-side database work, network delays or GPU rendering.

On a permitted development client, Resmon reports resource CPU and memory usage. It is a developer command, so a normal client may return an access-denied message. Follow the official console documentation rather than assuming the command is broken. (Cfx Documentation)

Our Resmon troubleshooting guide explains how to reproduce an expensive action and investigate it further. A low resource-monitor reading should not be treated as proof that a custom map is cheap to render.

For deeper investigation, use the official Cfx.re profiler to identify the relevant execution path. Measure the problem before and after a change; an idle screenshot is not a substitute for testing actual gameplay. (Cfx Documentation)

Frequently asked questions

Should I put every FiveM setting on the lowest option?

Not as a permanent solution. It is a useful diagnostic comparison, but afterward you should restore settings individually. Otherwise, you may be giving up image quality without addressing the cause of the slowdown.

Is High texture quality bad for FPS?

Not necessarily. In NVIDIA’s original GTA V testing, texture-quality changes had little performance impact when memory was sufficient. Limited VRAM changes that trade-off, so test the setting on the servers you use rather than choosing it from GPU model alone. (NVIDIA)

Why is only one FiveM server slow?

Treat that as a reason to investigate the server’s particular workload—not immediate proof that its hosting is bad. Compare locations and repeatable actions, then give the administrators enough detail to investigate client scripts, streamed assets or server-side delays.

Will these settings guarantee 144 FPS?

No. They are a starting configuration and a testing method. Your result depends on the system and the workload, and client-side resource execution can remain a limiting factor even after graphics quality is reduced. (Intel)

The setup worth keeping

Start with the balanced preset, test a representative route and change one setting at a time. Keep the visual improvements that your system can comfortably handle, and investigate repeatable stutters rather than hiding them behind a higher average FPS.

The best FiveM settings are the ones that hold up during your normal session—not the ones that produce the biggest number while you stand still.