Skip to content
Join our Discord for updates & support Discord

Why BeamNG Mods Tank Your FPS

You install a hyper-detailed supercar, spawn it, and your framerate falls off a cliff. The last car ran fine, the map hasn’t changed, and nothing on your PC is different. The mod is the only new variable, and BeamNG.drive hands you enough tools to pin down exactly which one is to blame and why.

Not all mods cost the same, and they don’t cost it the same way

BeamNG’s whole draw is soft-body physics: every vehicle is a mesh of nodes (mass points) connected by beams (springs), and the simulation solves that network thousands of times per second. A simple stock car might be a few hundred nodes; a meticulously modeled mod with a full drivetrain, deformable interior, and separate suspension linkages can climb into the thousands. That count is the single biggest reason one car chugs while another flies, and it has nothing to do with how good your GPU is.

The heavy offenders fall into three categories, and they hit different parts of your PC:

  • Node/beam count is a CPU cost. More nodes and beams means more physics math per timestep. It shows up as low framerate even when you’re parked and staring at a wall.
  • High-resolution textures are a VRAM and GPU cost. A mod that ships 4K skins, normal maps, and interior textures with no compression eats video memory fast, especially with several such vehicles spawned.
  • Dense maps are also a GPU cost. Large custom levels packed with meshes, props, grass, and complex lighting hammer draw calls and can blow past your VRAM budget.

A car and a map can both drop you to 30 FPS for opposite reasons, and the fixes are different. So the first real diagnostic step is figuring out which side is bottlenecked.

GPU-bound or physics-bound? Read the graph

BeamNG has a live performance readout. Open the app menu (Esc) and enable the performance/debug overlay, or bind it to a key in the controls. It shows your render framerate alongside the physics timing. The number that matters is how much time the physics simulation eats relative to your frame budget.

The quick mental model splits cleanly:

  • Physics/CPU-bound: FPS tanks the instant a heavy car spawns, gets worse with each additional vehicle or with traffic on, and doesn’t improve when you look at the sky or drop your graphics preset. The car is too expensive to simulate.
  • GPU-bound: FPS is fine pointed at empty ground but collapses on a dense scene, a detailed cockpit, or a prop-heavy area. Lowering resolution, texture quality, or dynamic reflections helps immediately. That’s visuals, not physics.

The fastest tell costs you two seconds: with the suspect car spawned, aim the camera straight up at the sky. If FPS jumps back to normal, you’re GPU-limited and the culprit is visual — textures, reflections, the map. If it stays low, the CPU is drowning in physics and the culprit is the car’s node-beam complexity. This one test replaces a lot of guessing.

Isolate the suspect on a clean stage

You can’t diagnose a mod on a busy scenario with traffic running — too many variables. Strip the environment down first.

  • Spawn on Gridmap (or the small, flat Automation Test Track). Gridmap has almost no environmental GPU cost, so whatever framerate you get is mostly the vehicle itself.
  • Turn traffic off and clear other vehicles so only your suspect is spawned. Every extra vehicle is another full soft-body simulation competing for CPU.
  • Spawn the stock equivalent — a default car like the ETK 800 or Ibishu Covet — on the same map and compare. If stock runs at 120 and the mod runs at 45 on the identical empty level, the mod is unambiguously the problem, not your hardware.
  • Note the number, then repeat with the mod’s config variants. Often only the “full detail” or race config is expensive, and a lighter config is perfectly fine.

Got a stack of mods installed? Use the in-game mod manager (Repository / Mods menu) to deactivate everything except the one you’re testing. Deactivating rather than deleting lets you flip mods back on one at a time — the classic binary-search hunt. Reactivate in groups: if a batch of five is clean, the culprit lives in the other batch.

Cut the cost without uninstalling

Sometimes you want to keep a heavy car. The physics load is fixed per vehicle, so you can’t dial down a single car’s node-beam fidelity — but you can control how many expensive things run at once:

  • Reduce the spawned vehicle count. Each AI or parked car is a full soft-body sim. Four traffic cars can cost more CPU than your player car.
  • Lower the traffic vehicle count if you run traffic, or turn traffic off entirely when you just want to drive.
  • Pick a lighter config of the same vehicle when one exists — a street build over a fully detailed race build.

Dropping your graphics preset or resolution only helps the GPU side; it does nothing for a physics-bound car. Don’t waste time lowering shadows to fix a node-beam bottleneck.

Why a huge mod folder slows you down before anything spawns

Installed-but-unspawned mods aren’t free. On launch, BeamNG scans and indexes every active mod in your mods folder (under your user directory, typically DocumentsBeamNG.drive[version]mods on Windows), and the vehicle/part selector enumerates all of them. A folder with 150 active mods means longer load times, a sluggish spawn menu, and more memory used before you’ve driven a meter. If reaching the main menu takes ages, prune what you don’t use — deactivate mods in the manager or move the .zip files out of the folder.

The culprit-hunting checklist

  • Spawn on Gridmap, traffic off, only the suspect vehicle present.
  • Note FPS, then aim the camera at the sky. A big jump means a GPU/visual problem; no change means a physics/CPU problem.
  • Spawn the stock equivalent on the same map and compare numbers.
  • Open the performance overlay and watch physics timing against render FPS to confirm the bottleneck.
  • With several mods installed, deactivate all but one and binary-search the offender.
  • Test the mod’s lighter configs before writing it off — the “full” build may be the only heavy one.
  • Decide: keep it with fewer spawned vehicles, or swap in a better-optimized alternative.

Once you know whether you’re fighting nodes or textures, the call is easy. A gorgeous 3,000-node show car is a great screenshot machine and a poor choice for a 20-car pileup; a lighter, well-optimized version of the same car can get you 90% of the look at half the cost. When you go hunting for a replacement, our BeamNG mods section and the mod comments often flag which builds are performance-friendly.

FAQ

How do I know if it’s my GPU or the car’s physics?

Spawn the car on an empty map and point the camera at the sky. If FPS recovers, you’re GPU-bound and the cause is visual — textures, reflections, or a dense map. If FPS stays low no matter where you look, the car’s node-beam count is overloading your CPU, and no graphics setting will fix it.

Does having lots of mods installed hurt FPS even if I don’t spawn them?

Not your in-game framerate directly — unspawned mods don’t run physics. But a large active mod folder lengthens load times, bloats the vehicle selector, and uses extra memory at startup. If loading and menus feel slow, thin out the mods you don’t use.

Will lowering my graphics settings fix a car that lags when parked?

No. Lag while parked and looking at nothing is a physics/CPU bottleneck from the car’s node-beam count. Graphics settings only touch the GPU side. Reduce the number of spawned vehicles, turn off traffic, or switch to a lighter config of the car instead.