Collider types and performance

Optimize CPU performance by selecting the most appropriate colliderAn invisible shape that is used to handle physical collisions for an object. A collider doesn’t need to be exactly the same shape as the object’s mesh - a rough approximation is often more efficient and indistinguishable in gameplay. More info
See in Glossary type for your sceneA Scene contains the environments and menus of your game. Think of each unique Scene file as a unique level. In each Scene, you place your environments, obstacles, and decorations, essentially designing and building your game in pieces. More info
See in Glossary.

Choosing the right collider type directly affects CPU performance. Unity has several collider types, each with different computational costs.

To learn more about MeshThe main graphics primitive of Unity. Meshes make up a large part of your 3D worlds. Unity supports triangulated or Quadrangulated polygon meshes. Nurbs, Nurms, Subdiv surfaces must be converted to polygons. More info
See in Glossary Collider component properties, refer to Mesh collider component reference.

To learn more about collider interactions, refer to Collider interactions.

The following table summarizes the collider types and their performance characteristics in order from most to least performant:

Collider type	Performance characteristics
Sphere ColliderA sphere-shaped collider component that handles collisions for GameObjects like balls or other things that can be roughly approximated as a sphere for the purposes of physics. More info See in Glossary	The simplest and most efficient collider. Use for round objects and general-purpose interactions.
Capsule ColliderA capsule-shaped collider component that handles collisions for GameObjects like barrels and character limbs. More info See in Glossary	Slightly more complex than a Sphere Collider, but still efficient. Use for characters, poles, or other elongated shapes.
Box ColliderA cube-shaped collider component that handles collisions for GameObjects like dice and ice cubes. More info See in Glossary	Efficient and flexible, especially for rectangular or block-shaped objects. Slightly more resource-intensive than Sphere Collider or Capsule Collider.
Convex Mesh Collider	More resource-intensive than primitive colliders. Use only when primitive shapes or compound colliders cannot approximate the geometry. The mesh must be convex. You can attach this to GameObjectsThe fundamental object in Unity scenes, which can represent characters, props, scenery, cameras, waypoints, and more. A GameObject’s functionality is defined by the Components attached to it. More info See in Glossary with non-kinematic RigidbodyA component that allows a GameObject to be affected by simulated gravity and other forces. More info See in Glossary components attached.
Non-convex Mesh Collider	A Mesh Collider with Convex unchecked. The most resource-intensive collider type. Use only for static, non-moving geometry that requires precise collisionA collision occurs when the physics engine detects that the colliders of two GameObjects make contact or overlap, when at least one has a Rigidbody component and is in motion. More info See in Glossary surfaces. Cannot be attached to non-kinematic Rigidbodies.

Compound colliders

Compound colliders are formed by parenting multiple primitive colliders, such as Sphere, Box, and Capsule colliders, to a single GameObject with a Rigidbody component. You can use compound colliders, which are comprised of simple components, to approximate complex shapes. Compound colliders are generally more efficient than using a single, complex Mesh Collider for a dynamic object, but they’re more resource-intensive than a single primitive collider.

Use as few primitive colliders as possible within a compound setup to have effective collision detectionAn automatic process performed by Unity which determines whether a moving GameObject with a Rigidbody and collider component has come into contact with any other colliders. More info
See in Glossary.

To learn more about compound colliders, refer to Compound colliders.

Use Compound colliders when:

You have an object whose shape cannot be accurately represented by a single primitive collider.
You have dynamic objects that require a complex collision shape but must remain performant (as an alternative to a costly convex Mesh Collider).
You need dynamic concave collision shapes, which a single Mesh Collider on a dynamic Rigidbody cannot do because a single Mesh Collider must be convex.
You need fine-tuned collision zones for gameplay. For example you have distinct head, torso, or limb hitboxes on a character.

Prepare meshes for mesh colliders

Mesh Collider components require a preprocessing step called “mesh cooking” to convert their geometry into an optimized format for efficient physics calculations. If this cooking occurs at runtime, it can cause significant CPU performance spikes.

To learn more about Mesh Collider components, refer to Mesh collidersA free-form collider component which accepts a mesh reference to define its collision surface shape. More info
See in Glossary.

You can avoid mesh cooking at runtime by doing the following:

Use Editor-Time cooking (Implicit): The primary way to pre-cook meshes is in the Editor. When you assign a Mesh asset to a Mesh Collider component in the InspectorA Unity window that displays information about the currently selected GameObject, asset or project settings, allowing you to inspect and edit the values. More info
See in Glossary window and save your scene or prefabAn asset type that allows you to store a GameObject complete with components and properties. The prefab acts as a template from which you can create new object instances in the scene. More info
See in Glossary, the Editor automatically cooks the mesh for that specific collider instance. The cooked data is then saved with your asset.
Enable Prebake Collision Meshes in the Player section of the Project SettingsA broad collection of settings which allow you to configure how Physics, Audio, Networking, Graphics, Input and many other areas of your project behave. More info
See in Glossary: Enable Prebake Collision Meshes to make the Editor proactively find and cook all meshes used by Mesh Collider components in your project during builds, asset imports, and Editor loads. This setting is enabled by default. Prebake Collision Meshes can increase build times and Editor load times as the meshes process. The cooked data is stored, which slightly affects build size if many unique meshes are involved, but the main trade-off is increased build and load duration versus runtime stability. Use Prebake Collision Meshes to prevent unexpected runtime cooking spikes by ensuring most meshes are ready. The recommended best practice is to keep Prebake Collision Meshes enabled as a safety net. Consider disabling Prebake Collision Meshes only if you manually and meticulously manage all mesh cooking. For example, you rely solely on Editor-time instance cooking and Physics.BakeMesh for all dynamic cases and you want to potentially reduce build and Editor processing times.
Use Physics.BakeMesh: For meshes assigned or procedurally generated at runtime, use Physics.BakeMesh(meshInstanceId, convex, cookingOptions) to explicitly control when cooking occurs. Preferably use Physics.BakeMesh during loading screens or in background threads if you use the job system.

Additional resources

Select a broad phase pruning algorithm

Configure Mesh Collider component cooking options for optimization