How to Print Miniatures on an FDM Printer

Search “best FDM printer for miniatures” and you will find top-five lists. Most of them are honest in the way most gear lists are honest: they compare printers against each other, rank them by specs, and leave you with a recommendation. What almost none of them say is that the honest answer to “best FDM printer for miniatures” is mostly “buy a resin printer instead, here is the guide.” A $149 Elegoo Mars 4 prints miniatures that no FDM machine at any price point will match on surface finish.

That said, there is a working method for printing miniatures on FDM, and this is the article for people who already own an FDM machine and want to use it. Not because it is better than resin. Because it is the printer they have. The method is not what most YouTube tutorials suggest. The trick is to stop printing miniatures the FDM-traditional way and start printing them the resin way. That one change closes most of the gap for chunky models at tabletop viewing distance.

For the broader question of which technology to buy for miniatures, resin or FDM for tabletop terrain covers the trade-offs with the wider context. This article is the practical guide once you have decided FDM is the constraint you are working with.

Why the FDM-traditional method fails for miniatures

The standard FDM approach to printing any model is to lay it flat on the build plate, auto-generate support structures rising vertically from the plate, and print it at whatever layer height the preset offers. For terrain, boxes, and functional parts, this works perfectly well. For miniatures, it is the worst possible way to do it.

Here is why. A miniature printed flat on the bed presents every important vertical surface at the worst possible angle for layer line visibility. The face, the cloak folds, the weapon shaft, the chest armour: all of these surfaces stack horizontal layers directly perpendicular to the viewer. At that angle, every layer line is fully visible. You see the staircase.

Resin printing avoids this by tilting every model at 45 to 60 degrees on the build plate. The geometry that comes off a resin printer looks better not just because resin pixels are smaller, though they are. It looks better because the tilt puts visible surfaces at an angle where the layer line pattern is far less visible to the eye. Tilting the model distributes the layer lines across the surface rather than stacking them directly at the viewer.

This is exactly the principle to apply to FDM. Stop thinking of the model as something that sits flat and print it the way a resin printer would orient it.

The hardware you actually need

The technique only works with the right tooling. On the wrong hardware it will waste your time.

A 0.2 mm nozzle. This is non-negotiable. A 0.4 mm nozzle produces line widths that are simply too coarse for miniature geometry. At 0.2 mm, the line width drops to a level where a 32 mm figure starts to have readable surface definition. The Bambu A1 and the P1S both support hardened steel nozzles at 0.2 mm. A genuine hardened steel 0.2 mm nozzle for the Bambu system costs roughly twenty to forty dollars depending on the source, and it handles abrasive filaments like PLA-CF without wearing out. If your machine does not support a 0.2 mm nozzle natively, this technique is not available to you on that machine.

Fine layer heights. 0.08 mm for showcase quality, 0.12 mm for tabletop standard. These are slow. A 40 mm infantry figure at 0.08 mm takes several hours. That is the cost. At 0.12 mm the speed is more manageable and the quality is still well above what flat orientation at 0.2 mm produces.

Tree supports. Not the default slab supports. Not vertical pillar supports. Tree supports that branch from a thin root structure and contact the model at small, targeted points. OrcaSlicer’s organic tree supports are noticeably better than Bambu Studio’s tree supports for miniature work, and worth using if you can set them up.

PLA or PLA-CF. PLA is the sensible default. PLA-CF (carbon fibre reinforced PLA) carries a small advantage on miniatures because the carbon fibre stiffens small features against the stress of the print process, which reduces warping on thin protrusions like sword blades and antennae. The textural difference on the surface is negligible. For filament choices and what makes PLA-CF worth using, the filament guide for 3D printed terrain covers the material options in more detail.

A direct drive extruder helps but is not required. A Bowden setup can run this technique but requires more careful retraction tuning to avoid stringing on tree support branches.

The slicer setup

This is the section that earns the article its place. The settings below are tuned for miniature work, not terrain. They are also tuned for Bambu Studio and OrcaSlicer. If you are running a different slicer, translate the parameters across rather than looking for menu-identical options.

Orientation. Tilt the model 45 to 60 degrees on the build plate. The tilt axis depends on the figure. The usual approach for a humanoid miniature is to pitch the model forward so the face and chest tilt toward the plate, with the back of the figure elevated and presenting to the supports. This puts the surfaces the viewer reads first (face, chest) at the optimal layer line angle, and hides the support contact marks on the back.

Tree support settings. In OrcaSlicer, organic tree supports tuned for miniatures: support contact distance 0.10 to 0.15 mm, tree branch diameter 1.5 to 2 mm, tree top diameter 0.4 to 0.6 mm. These settings give thin, targeted support contacts that remove cleanly and leave minimal scarring. In Bambu Studio, set tree supports and reduce the contact layer Z distance to the minimum the slicer allows.

Layer height. 0.08 mm for a piece that will sit on a shelf and be examined at close range. 0.12 mm for a gaming table figure that reads at roughly a metre of viewing distance. Below 0.08 mm yields diminishing returns on a 0.2 mm nozzle; the line width becomes the limiting factor rather than the layer height.

Walls. Four to five perimeters. At miniature scale the walls overlap completely across most of the model’s geometry. Infill is effectively irrelevant; 100 percent infill is not wasted material here because the model volume is small and the walls define the surface you care about.

Speed. Outer wall at 60 to 80 mm/s. Inner wall and infill can run faster. Travel speed at 200 mm/s and above is fine because the travel distances on a small model are short. The risk of ghosting and ringing on a miniature at these dimensions is low enough that outer wall speed is the only speed parameter worth tuning carefully.

Z-seam placement. Set to random for organic shapes without a clear back surface (monsters, creatures). Set to aligned-to-back for humanoid figures where there is a defined back surface that faces away from the viewer. Never set to aligned-to-corner on a model without corners, which is most organic miniatures; this produces a visible vertical scar on the surface where the seam snaps to the nearest geometric corner it can find.

Pressure advance and coasting. Bambu firmware handles pressure advance automatically. Coasting helps clean up seam blobs; a value of 0.1 to 0.2 mm works on most setups but tune it on a test print.

First layer. 30 mm/s on the first layer, 105 to 110 percent flow, PEI plate or glue stick adhesion. Tree supports need a firm first layer bond or the whole structure peels mid-print. This is the most common failure mode on the first attempt at this technique; if the support base peels, slow the first layer and increase the flow before adjusting anything else.

The workflow

Print one test model first. Pick a figure you do not care about and run it at your chosen settings before committing to anything larger or more important. Tune until single-print quality is at the level you want. Only then batch.

The instinct with FDM is to fill a plate and run overnight. With miniatures at this scale and orientation that instinct is expensive. Eight models on the plate with support settings that are slightly wrong produces eight ruined prints across six hours of machine time. A single-print iteration loop is faster overall even though it feels slower.

What this is good for

Chunky 32 mm scale models that the viewer reads mostly through silhouette and form: Space Marines, Stormcast Eternals, heavily armoured dwarves, large-scale monsters, creatures with broad surface area. Vehicles with readable surface geometry. Terrain crossover pieces such as fortifications with integral mounted figures, or scenic bases where the figure and the base print together.

The technique also makes sense when you want multi-colour output through an AMS system, which is impossible on a resin printer. A Space Marine in chapter colours, a basing texture in a different filament: these are things FDM can do that resin cannot, and at this scale the base print quality is good enough for the multi-colour effect to carry the piece.

For heavy-coverage painting styles, drybrushing, Contrast paints, oil washes, the technique closes more of the gap than it does for display painting. Fine surface texture disappears under these treatments and a model at tabletop distance looks better than the slicer preview suggests.

What this is not good for

28 mm and smaller faces. Fine cloth folds. Intricate weapons: sword guards, filigree scrollwork, chains. Single-figure showpieces that will sit on a shelf and be examined at 30 centimetres. Any model where the viewer is going to look at the surface up close and the surface is the point.

For any of those, resin still wins decisively. An $149 Mars 4 at 18 micron pixels and 30 micron layers produces a surface finish that no FDM machine at any nozzle size matches on fine detail. If miniature painting is your primary craft and the figure matters, the resin buying guide is the right next read.

When FDM with this technique beats resin

When you already own a P1S or an A1 and the viewing distance is a metre or more. When your painting style is heavy coverage that masks fine surface texture. When you want batch printability across a range of sizes without the resin handling, ventilation, and cleanup overhead. When you are printing terrain and figures for the same game and want one machine handling both. When multi-colour output through filament switching is part of the plan.

The P1P is mentioned frequently in the Bambu ecosystem and it is a capable machine, but for miniature work the A1 and P1S are the better recommendations because the P1S in particular gives you the full enclosure for PLA-CF which is useful for this application. The P1P review for tabletop work covers where the P1P sits in the Bambu range if you are already deciding between the three.

When resin still wins

The surface quality argument is not close. Resin at 18 to 24 micron pixel size and 30 micron layer height produces detail that FDM at 0.2 mm nozzle and 0.08 mm layer height cannot match. The resin print looks like it was cast. The FDM print looks like it was printed.

For anyone focused on miniature painting as a craft, the right answer is still a $149 Elegoo Mars 4. No technique in this article changes that. The resin-style FDM method closes the gap. It does not erase it. The full resin buying guide covers what resin printing costs and what to expect from it.

The bottom line

The right printer for miniatures is still resin. But if FDM is the constraint you have, whether because you already own one, because resin handling is not viable in your space, or because you want one machine for both terrain and figures, the resin-orientation-and-support technique closes most of the gap for chunky models at tabletop quality.

The FDM-traditional method does not. Flat orientation, standard layer heights, vertical supports from the plate: that approach produces results that look printed from across the room. The resin-style method, 45 degree tilt, 0.08 to 0.12 mm layers, organic tree supports, takes more setup but produces something you can paint and put on a table without apologising for it.

For the hybrid route, where you print FDM bases and resin detail pieces together, FDM base, resin top: splitting big models covers the split-plane method and how the two materials join.