# Chamfer vs fillet in 3D printing: when to round an edge and when to cut it

When to fillet a 3D printed edge, when to chamfer it, the bottom chamfer that beats elephant foot, and the inside-corner fillet that stops parts cracking.

---

You printed a bracket, bolted it up, and it snapped clean off at the inside corner the
first time it took any load. Or you printed a little box and it rocks on the desk because
the bottom edges spread out into a lip. Or you rounded the bottom edge in CAD to make it
look nice, and the slicer buried the whole underside in supports.

Those are three different problems with one root cause: the edge treatment was wrong, or
in the wrong place. Every edge on a printed part is a decision. You can leave it sharp,
round it off with a fillet, or slice it at an angle with a chamfer. Get that decision
right and parts print cleaner, sit flat, and stop cracking where they used to. This post
covers the difference, when to reach for each, the exact starting numbers, and the one
rule that covers most parts.

## What is the difference between a chamfer and a fillet?

A chamfer is a flat, angled cut across an edge, almost always at 45 degrees. A fillet is a
rounded blend that replaces the sharp corner with a smooth arc. Both do the same basic job
of killing a sharp 90-degree edge, but the shape is different, and on an FDM printer that
difference decides two things that matter: whether the edge prints cleanly without
supports, and whether the part survives being loaded.

Think of it this way. A chamfer is what a file or a hand plane leaves. A fillet is what a
router bit or a blob of weld leaves. Neither is "better" in the abstract. They are tools
for different jobs, and a well-designed part usually has both.

## When should you use a fillet?

Reach for a fillet when the edge carries load, and above all on an inside corner. A sharp
inside corner is a stress concentrator: force piles up at that single point and the part
cracks there first. Rounding the corner spreads that force over an arc instead of a point,
and the strength gain is not small. On a typical bracket, filleting the inside corner can
more than double how much load it takes before it fails.

Inside corners matter even more on FDM prints than on machined parts, because printed
parts are weakest along their layer lines. A sharp internal corner combines a stress
concentration with that layer-line weakness in exactly the same spot, which is why a
poorly filleted bracket almost always breaks at the base of the vertical wall. A fillet
there gives the load somewhere to flow.

How big? A good starting rule is a fillet radius at least equal to the thickness of the
wall next to it, and usually in the 1 to 3 mm range for functional parts. Bigger helps
strength but eats into the part; a radius roughly matching the wall is the sweet spot.
That relationship to wall thickness is worth understanding on its own, and it ties
directly into
[how thick your printed walls should be](/blog/how-thick-should-3d-printed-walls-be).

One catch: a fillet on an *outside* bottom edge, the edge that sits down toward the bed,
curves under itself and becomes an overhang. That is a printability cost you take on for a
bit of cosmetics or strength, and it is the whole reason the chamfer exists as an
alternative.

## When should you use a chamfer?

Reach for a chamfer when the edge needs to print cleanly, guide an assembly together, or
sit flat on the bed. Because a chamfer is a straight angled face rather than a curve, it
does not curl under into a nasty overhang the way a bottom fillet does, so it prints crisp
with no support in places a fillet would struggle.

The three everyday jobs for a chamfer:

- **Lead-ins.** A small chamfer on the end of a pin, peg, or the mouth of a hole acts like
  a funnel, so two parts start square and slide together instead of catching on a sharp
  lip. A 0.5 to 1 mm chamfer is plenty.
- **Bottom outside edges.** A chamfer on the edge that meets the bed both hides the first-
  layer bulge (more on that next) and gives a cleaner base than a sharp edge that splays.
- **The top of a horizontal hole.** Chamfering the top lip of a hole that prints sideways
  shortens the unsupported bridge across it, so the top of the hole comes out round
  instead of sagging.

Chamfers are also just easier to model and easier to print. When you are not sure which to
use and the edge is not a load-bearing inside corner, a chamfer is the safe default.

## How do you fix elephant's foot with a bottom chamfer?

Add a small 45-degree chamfer, roughly 0.4 to 0.8 mm, to the outside bottom edges of the
part. Elephant's foot is that squished-out lip on the first layer or two, caused by the
weight of the part and a hot bed pressing the bottom layers wider than the rest. A bottom
chamfer beats it by making each of those first few layers slightly smaller than the one
above, so as the bottom spreads outward it is only filling in the angle you cut, not
bulging past the true edge of the part. The part comes off the bed with a clean, square
base instead of a lip you have to sand.

This is a design-side fix, and it stacks with the slicer-side one. Most slicers have an
"elephant foot compensation" setting that shrinks the first layer in software. That works
too, but it lives in your slicer profile, so it travels with your machine, not with the
model. A modeled chamfer is baked into the part itself, so anyone who prints your file
gets a clean base on any printer. For a light case, 0.4 mm is enough; for a stubborn one,
go to 0.8 mm or more.

## Chamfer vs fillet: quick comparison

| | Chamfer | Fillet |
|---|---|---|
| Shape | Flat angled cut (usually 45°) | Rounded blend |
| Best at | Printability, lead-ins, flat bases | Strength, stress relief, cosmetics |
| On a bottom outside edge | Prints clean, fights elephant foot | Curls into an overhang, may need support |
| On a loaded inside corner | Helps a little | Best choice, spreads the stress |
| Ease to model and print | Simplest, most forgiving | Slightly more overhang-prone |

## What size chamfer or fillet should you use?

Numbers beat adjectives. Here are sane starting points by where the edge lives. Treat them
as a first print, then adjust.

| Where the edge is | Use | Starting size |
|---|---|---|
| Bottom outside edge (sits on the bed) | Chamfer 45° | 0.4 to 0.8 mm |
| End of a pin or mouth of a hole (assembly lead-in) | Chamfer | 0.5 to 1 mm |
| Top of a hole that prints sideways (bridge) | Chamfer | 0.5 to 1 mm |
| Inside corner under load (bracket web) | Fillet | At least wall thickness, ~1 to 3 mm |
| Outside vertical edge (looks, handling) | Either | 0.5 to 2 mm |

A fillet on a bottom outside edge is the one combination to think twice about. It looks
great in CAD and turns into an overhang on the printer. If you want that rounded base and
you do not want the supports, either flip the part and print it the other way up, or accept
a chamfer there instead. The tradeoff is the same one covered in
[overhangs and supports](/blog/overhangs-and-supports).

## The one rule that covers most parts

If you remember nothing else: chamfer the outside bottom, fillet the inside corners.

That single habit handles the two most common failures at once. The bottom chamfer keeps
the base clean and square and dodges elephant's foot with no supports. The inside-corner
fillet stops the part cracking where the load piles up. Everything else, the cosmetic
outside edges and the assembly lead-ins, is a smaller call you can make by feel. Get the
bottom and the inside corners right and most functional prints come out both stronger and
cleaner.

<TryMeshra />

## How Meshra puts chamfers and fillets on sliders

Here is where designing the part yourself, instead of downloading a finished mesh you
cannot edit, pays off. In Meshra you describe the part in plain English, including the edge
treatment, and it writes real parametric CAD (CadQuery on the OpenCascade kernel) and
builds an exact solid, not a triangle mesh. You could ask for "an L-bracket with a 2 mm
fillet on the inside corner and a 0.6 mm chamfer on the bottom edges," and get a solid part
with those edges actually modeled where you asked.

The payoff shows up on the second print. When a dimension like that chamfer or fillet size
is one of the parameters Meshra generates for the part, it becomes a slider in the
parameter drawer. If the elephant's foot is still poking out, drag the bottom chamfer from
0.6 up to 0.8 mm and the same code re-executes deterministically with the new number: the
chamfer grows, everything else stays exactly where it was, there is no AI call, and it is
instant and free on every plan. Widening an inside fillet to add strength is the same one-
slider change instead of a remodel.

Meshra also gives you a sanity check before you slice. The printability X-ray and its
plain-language summary read the loaded part and flag overhanging faces (for example
"3 faces may need supports") and walls that may be too thin. That is exactly what catches a
bottom fillet you left in by accident: the summary shows it as an overhang, so you can
swap it for a chamfer before you waste a print. When the part is right, export STL for your
slicer, or STEP if you want to open those edges in another CAD tool later. For more on how
the slider-driven, deterministic re-execution works,
see [from a sentence to a printable part](/blog/from-a-sentence-to-a-printable-part).

## FAQ

### Is a chamfer or a fillet stronger?

A fillet, on a loaded inside corner. It replaces the sharp point where stress concentrates
with a smooth arc that spreads the load, and on a typical bracket that can more than double
the force the part takes before it cracks. A chamfer helps a little but not nearly as much.
For pure printability or a lead-in, the strength difference does not matter and the chamfer
is the easier choice.

### Should I chamfer or fillet the bottom edge of a print?

Chamfer it. The bottom outside edge sits against the bed, and a fillet there curls under
into an overhang that needs support, while a 45-degree chamfer prints clean and also fights
elephant's foot. Save fillets for inside corners and for outside edges that are not facing
down.

### How big should a fillet be?

Start with a radius at least equal to the wall thickness next to it, usually 1 to 3 mm for
functional parts. Bigger spreads stress better but removes more material; a radius that
roughly matches the wall is a good balance. For a purely cosmetic edge, anything from 0.5
to 2 mm reads as "finished" without much thought.

### Will a fillet on the bottom of my part need supports?

Usually yes, if it is on an outside edge facing the bed. That rounded surface starts nearly
horizontal where it meets the bed and curves up, which is exactly the geometry a printer
struggles to lay down in mid-air. A chamfer in the same spot avoids the problem entirely.
If you must have the rounded look, reorient the part so that edge is not on the bottom.

### Can I add chamfers and fillets without learning CAD?

Yes. That is the point of describing the part in plain language: you say "0.6 mm chamfer on
the bottom edges" and Meshra models it. If the fillet or chamfer size is one of the
dimensions it generated, you then fine-tune it with a slider, no CAD software and no AI call
for the tweak. It is the fast way to iterate on an edge that came out slightly wrong.

## Where to start

Pick a part you have already printed that cracked at a corner or came off the bed with a
lip, and rebuild it in the [Meshra builder](/templates) with the edges called out:
fillet the inside corner, chamfer the bottom. Print it once, then tune the numbers with a
slider from what you see. The [template gallery](/templates) has brackets, mounts, and
enclosures with parameters already exposed if you would rather start from a working part
than a blank prompt, and your first part is free with no card. See [pricing](/pricing) for
what each plan includes, or [create an account](/sign-up) to keep your parts and dial in
the edges after the first print.

Sharp corners are not a rule you have to live with. Round the ones that carry load, cut the
ones that sit on the bed, and when the first print is close but not perfect, change the one
number that is off instead of starting over.