Lightning Infill: How It Works and When to Use It
Lightning infill is the fastest infill pattern in modern slicers — but it's not a universal default. A breakdown of how it differs from gyroid, cubic, and grid, and where it's the right choice.
Lightning infill is a relatively new pattern — first added to Cura around 2021 and propagated into PrusaSlicer (2.5+), OrcaSlicer, and Bambu Studio since then — and it does something none of the older patterns do: it generates infill only where the slicer thinks the top layers need support, then stops.
The result is a print that uses a fraction of the filament and prints a fraction of the time, with a top surface that’s still flat. The catch is that the part has very little strength below the top surface. That’s fine for some prints and disastrous for others.
What Lightning actually does
Conventional infill patterns (grid, gyroid, cubic, triangle) fill the entire interior of a part with a regular structure at the percentage you specify. A 100 mm cube at 15% gyroid has gyroid throughout, top to bottom.
Lightning is different. It starts at the top surface of every flat top region and grows a tree of supports downward — branching, spreading, and reducing — only as much as needed to keep the top surface from sagging. Below those tree-of-support areas, the part is mostly hollow.
In a vertical cross-section of a Lightning-filled part:
- Solid top shell at the top.
- A tree of thin branching infill lines reaching upward from below to support that shell.
- Empty space everywhere the branches don’t reach.
- Solid bottom shell at the bottom.
Side walls (perimeters) carry all the load between top and bottom.
When Lightning is the right choice
Display models / showpieces
If the print is for visual display only — a figurine, a desk decoration, a model car, a vase, a planter — it doesn’t need internal strength. Lightning saves 30–50% on filament and 20–40% on print time vs. gyroid at 15%, with no visible quality difference on the outside.
Drafts and prototypes
Quick iteration prints where you’re checking fit, shape, or proportions don’t need full strength. Lightning gets you a part to inspect faster.
Prints where appearance matters more than strength
Cosplay props, sculptural pieces, replicas, anything that will be handled gently and shown off — Lightning is the right default.
Prints where you’ll add reinforcement separately
If you’re printing a shell that will be filled with epoxy or polyurethane after printing (a common technique for large display pieces), the part doesn’t need printed infill for strength. Lightning works fine and the post-print fill carries the load.
When Lightning is the wrong choice
Mechanical parts
Brackets, gears, mounts, anything load-bearing — Lightning is dangerous. The part has near-zero shear strength in the Z direction and minimal compressive strength anywhere the branches don’t reach. A 15% gyroid bracket can carry significant load. A 15% Lightning bracket will fail under the same load.
Parts that will be drilled or tapped post-print
Drilling into a Lightning-filled part means drilling into mostly air. Threads cut into the bottom of a Lightning print will have nothing to grip below the top shell.
Parts that need to flex without breaking
Lightning’s branching structure has stress concentrators where branches meet the top shell. Repeated flexing fractures at these points. Use gyroid for any part that needs to flex.
Parts where the bottom surface is also a “top” — bowls, cups, inverted shapes
Lightning generates from the top down. If your part has a bottom that’s structurally a top (a cup interior, a bowl base), Lightning won’t support that surface from below. You’ll get sag or holes in surfaces the slicer didn’t recognize as “top.” This is a common gotcha.
Vase mode
Vase mode uses no infill at all, so the question is moot. But if you’re transitioning between vase-mode walls and conventional walls in the same print, Lightning won’t carry through cleanly. Use grid or gyroid for the conventional sections.
Lightning vs. the alternatives at low percentages
A useful way to think about Lightning is “what infill pattern works best at 5% density?” because that’s where Lightning’s sparseness lands.
| Pattern | 5% behavior | 15% behavior | Strength at 15% |
|---|---|---|---|
| Lightning | Top-supporting tree | Top-supporting tree, slightly denser | Low |
| Gyroid | Sparse 3D coils | Decent 3D coils | High |
| Cubic | Sparse cube struts | Open cubic lattice | Medium-high |
| Grid | Two-direction parallel lines | Parallel lines, all at one angle | Low (only strong on Z plane) |
| Triangle | Sparse triangular cells | Triangular lattice | High |
At 5% density, Lightning prints faster than any of the others and produces a flat top surface. Gyroid at 5% looks fine on the outside but is structurally weaker than gyroid at 15% — it offers no advantage over Lightning for visual prints.
Above 15% density, Lightning’s time savings shrink because the supports get dense enough that the time difference between “tree of supports” and “full gyroid” narrows.
Settings to dial in Lightning
Infill density
Lightning’s density slider behaves differently from other patterns. At 0% it generates almost nothing (top will sag). At 5–10% it generates a usable tree. Above 15% the savings shrink rapidly. The sweet spot for most display prints is 8–12%.
Top shell layers
Because Lightning supports only the top shell, the top shell needs to be thick enough to bridge between branches. Default top shell layer counts (3–5) are usually fine, but on prints with widely-spaced Lightning branches you may need 5–7 top layers to prevent visible pillowing.
Rule of thumb: top thickness should be at least 1.0 mm. At 0.2 mm layers that’s 5 layers; at 0.12 mm layers that’s 9 layers.
Lightning support angle
PrusaSlicer exposes a “Lightning support angle” parameter that controls how much the branches can lean. Default 40° works for most prints. Increase to 50–55° if you’re seeing too many branches; decrease to 30° if the top surface is showing pillowing.
Bambu Studio’s equivalent (Tree support branch angle, when Lightning is selected) defaults to a similar value.
Combined with thicker layers
Lightning pairs well with thicker layer heights (0.28 mm on a 0.4 mm nozzle). The thicker layers fill in the gaps between branches faster, and the time savings compound — a 0.28 mm Lightning print can be 60% faster than a 0.2 mm gyroid print on a large display piece.
A quick recommendation tree
- Display piece, no mechanical load → Lightning at 10%.
- Functional bracket → Gyroid at 20–25%.
- Flexing part (clip, hinge) → Gyroid at 25–30%.
- Prototype to check fit/scale → Lightning at 8%.
- Part you’ll drill/tap later → Gyroid at 30% in the drilled regions, Lightning elsewhere (use modifier meshes for this).
Cross-references
For the related question of perimeter strength and how it interacts with infill choice, see our Arachne perimeter generator guide. For mechanical-strength testing methodology on different infill patterns, PrintLabGuide ↗ is the next resource.
Related
Seam Placement: How to Hide the Z-Seam on FDM Prints
Every FDM print has a seam where each layer starts and ends. A practical breakdown of the seam-placement options in PrusaSlicer, Bambu Studio, OrcaSlicer, and Cura — and which one you actually want.
Adaptive Layer Height: When It Helps, When It Hurts
Adaptive layer height varies layer thickness across your print to save time on vertical sections without sacrificing detail on shallow overhangs. Here's how it works and when it's worth enabling.
Tree Supports: How to Configure Them for Clean Removal
Tree supports save filament and remove cleanly when configured well — and leave a mess when they aren't. A practical guide to tip diameter, branch angle, and interface layers in PrusaSlicer, Bambu Studio, and OrcaSlicer.