In brief
- The Distribute Points on Faces node creates points on a mesh with faces.
- Random is described in the manual as the fastest method, while Poisson Disk accounts for a minimum distance between points.
- The Density value specifies the number of points to distribute per square meter of mesh surface.
- Instance on Points can use repeated geometry as references at each point without duplicating the underlying data.
Blender’s official manual describes the Distribute Points on Faces and Instance on Points nodes, which can be used together to first create points on a mesh surface and then place geometry at those points. This guide covers this foundational Geometry Nodes workflow and its key settings. Since the source documentation does not specify a particular Blender version, the explanation is based on node behavior described on the relevant Blender Manual pages.
1. Use a mesh with faces for distribution
The Mesh input of the Distribute Points on Faces node must be connected to a mesh that has faces. The node generates points across the surface, so geometry without faces is not suitable for this input. The Selection input determines which face corners are considered for point distribution. This makes it possible to limit where distribution occurs using geometry data.
The node transfers the input geometry’s point, corner, and polygon attributes to the generated points. These include vertex weights and UV maps. It also provides Normal and Rotation outputs. Normal gives the normal of the triangle containing the point; Rotation is an Euler rotation derived from that normal. These outputs can be used to orient instances in subsequent nodes.
The node also creates the built-in id attribute, a stable identifier for each point. According to the manual, when the mesh is deformed or the density changes, the IDs of points that remain present stay consistent. This ID can be used in nodes such as Random Value and Instance on Points. Therefore, id data becomes important when points need to be identified in later steps of the distribution.
2. Choose Random or Poisson Disk
The Distribution Method setting determines how points are distributed. Random places points randomly across the surface and is described in the documentation as the fastest option. If you do not need to maintain a specific distance between points, this method is a straightforward starting point.
Poisson Disk also distributes points randomly, but takes the minimum distance between points into account. With this method, the Distance Min setting determines the minimum distance allowed between points. When Distance Min is zero, the node behaves the same as the Random method; no points are excluded because of the distance between them.
| Setting | Suggested value | Effect |
|---|---|---|
| Distribution Method: Random | No fixed value is specified in the source | Distributes points randomly; described in the documentation as the fastest method. |
| Distribution Method: Poisson Disk | No fixed value is specified in the source | Takes the minimum distance between points into account. |
| Distance Min | A value of zero behaves the same as Random | Sets the minimum distance between points; available only with Poisson Disk. |
| Density | Number of points per square meter | Determines the point distribution density on each mesh surface. |
| Seed | No fixed value is specified in the source | The random value used to generate points. |
No numerical recommendations absent from the documentation have been added to this table. Values should be determined by testing for the needs of the scene; with Poisson Disk in particular, minimum distance and density should be considered together.
3. Control density and distribution limits together
The Density input specifies the number of points to distribute per square meter on each mesh face. This value is multiplied by the Density Attribute values. Poisson Disk also has a Density Max input, which defines point density per square meter and contributes, together with Density, to determining the final density. Density Max is available only in Poisson Disk mode.
Increasing density does not always add more points. The manual notes that Density Max can be limited by Distance Min. Once the density exceeds what the minimum distance allows, no new points are added. So if you cannot achieve a denser distribution, check Distance Min as well as Density.
Seed is the random seed used to generate points. Change it when you need a different distribution. If you want to preserve the current distribution or compare changes, check the Seed value too. The source documentation does not recommend a specific Seed value.
4. Instance geometry on points
Connect the Points output of Distribute Points on Faces to the Points input of Instance on Points. Connect the geometry to be duplicated to the Instance input. Instance on Points adds the geometry at each point as a reference, without duplicating the underlying data. The Blender manual describes this approach as a fast way to use the same geometry many times in a scene.
The Rotation input determines the rotation of each instance, while the Scale input determines its size. The Rotation output of Distribute Points on Faces can be connected to orient instances according to the surface normal. However, the Z axis may be undefined in the rotation derived from the normal, because the surface normal does not provide enough information to define all three rotation axes. So if instances need a specific alignment on every axis—not just alignment to the surface—check the result.
Instance on Points makes the points’ attributes available in the instance domain of the resulting instances. If the input geometry has an id attribute, that value is transferred to the output instances. If the point input already contains instances, the node can place new instances on the points within them; the manual describes this as nested instancing.
When the Instance input uses geometry made up of multiple instances, the Pick Instances option lets you select one instance from the list for each point instead of adding the entire list. The selection is made using Instance Index; this option is available only when Pick Instances is enabled. The manual associates this use with the Collection Info node. Negative indices and indices greater than the number of items in the list wrap around from the other end. Use the Make Instances Real operator when you need to convert instances into real objects.
What does this offer in architectural and visualization workflows?
This pair of nodes provides a procedural foundation for arranging repeated elements across a mesh surface and using the same geometry at many points. In architectural visualization, the choice of method depends on whether the arrangement should be random or needs a minimum spacing between points. Instancing makes it possible to repeat the same geometry without recreating its data at every point; however, the source documentation provides no performance measurements or specific hardware requirements.
The documentation does not specify which Blender versions support these nodes or provide a compatibility table. The source metadata for the manual page states a CC BY-SA 4.0 license; this is the license information for the official documentation content discussed here. The sources provide no separate version information for the Blender installation or project files used.
Common mistakes
Connecting geometry without faces to the Distribute Points on Faces node. The input must contain a mesh with faces.
Assuming that increasing Density in Poisson Disk will always produce more points. Distance Min can limit the allowed density.
Expecting points to be kept apart when Distance Min is zero. In this case, the behavior is the same as the Random method.
Assuming that the Rotation output derived from the normal precisely determines all axes. The Z axis may be undefined.
Assuming that instances are always real geometry. The manual specifies the Make Instances Real operator for converting them into real objects.
Sources
2 sourcesSource texts are not republished; short quotes are marked, everything else is our own summary and commentary.
For architecture offices and visualization teams in Turkey, the practical value of this workflow lies in its ability to place repeated elements across surfaces in a controlled way. Random offers a fast distribution option, while Poisson Disk is useful when spacing needs to be controlled. Using geometry as a reference also helps avoid recreating the same data at every point.
Density alone, however, does not explain the result: the Distance Min limit and the axis constraint in rotations derived from the normal should also be checked. The sources provide no hardware requirements or performance measurements. Although the official manual’s CC BY-SA 4.0 license is stated, this is not an explanation of Blender software’s licensing terms. No version information is provided either, so teams should verify the behavior in their own installations.
Frequently asked questions
Which node distributes points across a mesh surface in Blender?
The Distribute Points on Faces node creates points on a mesh with faces. Geometry without faces is not suitable for the node’s Mesh input.
What is the difference between Random and Poisson Disk distribution?
Random is described in the documentation as the fastest distribution method. Poisson Disk takes the minimum distance between points into account; when Distance Min is zero, it behaves the same as Random.
Does Instance on Points create real copies?
The node can use geometry as references at points, creating many instances in the scene without duplicating the same data. The manual also specifies the Make Instances Real operator for converting instances into objects.


