Infill Patterns Explained: Gyroid, Grid, Honeycomb and the Rest

Slicers ship a dozen infill patterns and explain almost none of them. The density slider decides how much plastic goes inside your part; the pattern decides where that plastic goes, and therefore which direction the part is strong in, how fast it prints, and whether the nozzle knocks its way through every layer.

Quick Comparison

Lines and Rectilinear

Parallel lines that alternate direction each layer. It is the cheapest, fastest pattern and produces a part that is stiff along the line direction and weak across it. Because each layer only crosses the one below rather than interlocking with it, vertical strength is poor.

Fine for prototypes and display pieces at low density. Not a choice for anything that gets handled.

Grid

Lines in two directions within the same layer, forming an even square lattice. Strong enough in the plane of the layer, fast, and predictable, which is why it’s a common default.

Its one real flaw is that the crossing points stack directly above each other every layer. The nozzle passes over its own intersections, and at higher densities or speeds this produces a ticking or knocking sound and can leave a slightly rough interior. Not a problem for most prints, occasionally a problem at 40%+.

Triangles and Tri-hexagon

Triangles are geometrically rigid, so this is the pattern to pick when the load pushes sideways across the part rather than down onto it. Tri-hexagon layers three line directions instead of two, spreading strength more evenly in plane.

Both cost more filament and print time than grid at the same density because there are more direction changes per layer.

Cubic and Adaptive Cubic

Cubic stacks tilted cubes so the pattern has structure in all three axes rather than just the layer plane. That makes it genuinely three-dimensional, and a good choice for functional parts that get loaded from unpredictable directions.

Adaptive cubic subdivides only near the part’s surfaces and leaves the deep interior sparse. On a large part it can cut print time substantially with little practical strength loss, because the interior of a big part contributes very little anyway.

Honeycomb

Hexagonal cells, the shape nature settles on for a reason: it is the most material-efficient way to fill a plane with a stiff structure. In-plane strength is excellent and it looks superb through a translucent print.

The cost is speed. Hexagons mean constant direction changes, and depending on the slicer, retractions where the path breaks. Expect noticeably longer prints than grid at the same density. “3D honeycomb” variants add cross-layer structure at a further speed penalty.

Gyroid

A continuous, smoothly curving surface that never intersects itself, printed as a wave that shifts phase each layer. It is the pattern this app is named after, and it earns the attention.

What makes it unusual:

• Isotropic. Strength is close to equal in all three directions, so you don’t have to predict the load direction to choose it well.
• No self-intersections. The nozzle never crosses its own path, so no knocking and no stacked intersection points.
• Continuous path. Fewer retractions than honeycomb, so it prints faster than its complexity suggests.
• Excellent for flexibles. TPU printed with gyroid infill compresses evenly instead of collapsing along a line direction, which is why it’s the standard choice for gaskets and squishy parts.
• Good for resin drainage and epoxy filling, if you do that sort of finishing.

The trade is that it’s slower than grid, and at very high densities the curves get expensive. Below roughly 40% it’s an excellent default for nearly any part.

Lightning

A tree-like structure that grows up from the bed only where it is needed to hold up the top surfaces, leaving the rest of the interior empty. It is the fastest and lightest option by a wide margin.

It provides essentially no strength. That’s not a flaw, it’s the design goal. For a large display model or a prop shell where the only job is supporting the roof, lightning can cut hours and hundreds of grams off a print.

Concentric

Rings that follow the outline of the part inward. It has no crossing structure, so as rigid infill it is weak, but for flexible filament it is excellent: the part compresses and springs back evenly rather than buckling along a lattice.

Also useful for parts where you want a clean, consistent look through a translucent shell.

Gyroid vs Grid vs Honeycomb

The three-way comparison people actually want:

• Speed: grid, then gyroid, then honeycomb.
• Filament at the same density: roughly equal; honeycomb slightly higher in practice because of path overhead.
• In-plane strength: honeycomb, then gyroid, then grid.
• Vertical and shear strength: gyroid clearly, because the others are essentially stacked 2D patterns.
• Print quality: gyroid and honeycomb avoid grid’s stacked-intersection knocking.
• Looks through a translucent wall: honeycomb, then gyroid.

If you want one answer: gyroid at 15-25% for almost everything, grid when you want speed and the part is decorative, honeycomb when the part is loaded flat and you don’t mind waiting, lightning when nothing but the top surface matters.

Matching Pattern to Job

• Miniature or bust - lightning or gyroid at 5-10%.
• Household part, organizer, toy - grid or gyroid at 15-20%.
• Bracket, enclosure, jig - gyroid or cubic at 25-40%, three or more walls.
• Gear, mount, tool body - gyroid or cubic at 40-60%, printed so load runs across layers.
• Gasket, grip, anything in TPU - gyroid or concentric at 10-20%.
• Large prop shell - lightning at 5-8%.

Pattern and density are one decision, not two. Pick the pattern for how the part is loaded, then set the density for how hard, and log the pair with the print so the combination that worked comes back next time.