What Are Perler Beads? Origins, Common Names, and Who They Suit

Perler beads are small plastic beads shaped like a short cylinder or a flattened disc, with a round hole running straight through the middle. You do not thread them. You stand each bead on the raised pegs of a pegboard, follow a pattern, then cover the board with ironing paper and heat it with an iron so that neighbouring beads fuse where they touch. Once it cools, the beads hold a fixed shape as one piece.

The material is usually a thermoplastic such as low-density polyethylene (LDPE), which is why the beads soften when heated and harden again as they cool. At room temperature the beads are hard and slightly brittle. Under heat their surfaces become tacky and stick to each other, but the central hole never closes on its own: the hole is flattened by the iron, not melted away. That single detail explains why the two faces of a finished piece look so different.

Fuse beads are not a beading or sewing craft. They behave more like physical pixel art: one bead is one pixel, and a pattern is a grid table of color codes. That is also why software can generate patterns automatically. Map each pixel of an image to the nearest available bead color and you have a pattern someone can actually build.

Fuse beads spread through Scandinavian and North American toy makers around the 1960s, and the Danish brand Hama and the American brand Perler became the two best-known names. Because each dominated a large market, both brand names turned into generic category words: in the US and UK people simply say perler beads or Hama beads, while Chinese speakers say 拼豆 or 拼拼豆豆.

The naming is genuinely confusing when you search for supplies. In English you will meet fuse beads, melty beads, iron beads, and the two brand names. In Japanese the product is アイロンビーズ, literally iron beads. In Chinese you also see 拼豆豆 and 豆豆画. All of these describe the same craft; only the mix of brand names and generic names differs.

One caution follows from this: saying I bought Hama, so my beads must be 5 mm is not reliable. Hama, like most manufacturers, produces more than one line, so the diameter printed on the packaging is what matters, not the brand name. Buying the wrong line means beads that will not sit on the board you already own, which is one of the most common and most easily avoided beginner mistakes.

These are brand names, not size standards. A single brand usually sells several lines, and while beads from different brands can share the same diameter, their color code systems almost never match. Perler is the common North American brand, Hama is Danish, Nabbi is another familiar Scandinavian brand, and IKEA Pyssla is IKEA's in-house line. Artkal and MARD are widely available through online craft channels and offer larger color ranges.

For a pattern generator, what gets modelled is the palette, not the brand. BeiYa supports nine brand palettes plus a 48-color universal palette, ranging from 18 colors for IKEA Pyssla to 221 colors for MARD. Choosing the right palette matters because the code printed on your pattern has to be a code that actually exists in the beads you own. Otherwise you stall at the first color you cannot find.

The differences that actually affect a project come down to three things: how many colors a brand offers, how it names them, and whether it sells special-effect beads such as pearl, glow, translucent, or glitter. The first two are visible directly in palette data. The third decides whether a cross-brand substitution is realistic, because a color-difference algorithm compares color values, not finishes.

Adding up the nine brand palettes gives 1,177 color cards. Removing duplicate codes leaves 988 unique color codes. The gap exists because different lines inside the same brand reuse one numbering scheme: several Artkal lines start with S or M, so identical-looking codes refer to different beads. That is also why the two figures have to be quoted separately rather than treated as one number.

The 48-color universal palette is deliberately excluded from both figures, because 21 of its codes already appear in the brand palettes and mixing them in would produce a number that is neither a brand total nor a full selectable total. You can still choose the universal palette in the generator; it simply is not counted in 1,177 or 988.

In practice those two numbers answer one question: can this tool match the beads I already own? With the 221-color MARD palette roughly a quarter of the 988 unique codes are selectable. With the 18-color IKEA Pyssla palette the range is narrow and the pattern will merge many colors into a few codes, so gradients will not survive. That is a consequence of palette size, not of the matching algorithm.

The entry barrier is low. You need no drawing skill, only the ability to count grid squares and place beads on the matching pegs with tweezers. That is why fuse beads work simultaneously as fine-motor practice for children, a stress-relieving hobby for adults, and a way for pixel artists to turn digital art into a physical object.

Choosing a size is mostly about age and purpose. Maxi 10 mm beads are large and easy to grip, which suits young children. Midi 5 mm is the mainstream grade and the best balance between detail and difficulty. Mini 2.6 mm allows the finest patterns, but a finished piece of the same physical size needs close to four times as many beads, and it demands noticeably more patience with tweezers.

Two safety rules apply everywhere. First, beads are small parts with a choking risk, so children under three should not handle them and any young child needs supervision. Second, ironing is an adult task: both the iron soleplate and a freshly heated piece are hot enough to burn. These rules do not depend on age recommendations; they apply to every fuse bead project.

A bead pattern is a quantised image. The original picture is cut into a grid, each cell takes one color, and each cell is then matched to the closest bead in the palette. A denser grid preserves more detail and produces a larger finished piece. Those two outcomes are tied together, so you cannot take one without the other.

Color mapping usually relies on the CIEDE2000 color-difference formula, which converts image pixels and candidate bead codes into CIELAB space and compares perceived difference, picking the code with the smallest ΔE00. Compared with a raw RGB distance, this tracks human judgement much better in skin tones, greys, and dark regions, and it is far less likely to declare two obviously different colors to be the same.

When an image has smooth gradients and the palette is small, banding appears. Floyd–Steinberg dithering addresses this by spreading the quantisation error of each cell into the neighbouring cells to the right and below, so adjacent beads alternate between two similar codes and the eye blends them at a distance. The cost is a busier surface, more colors to buy, and a pattern that is easier to misread while building.

The first is treating fuse beads as a kind of beading. Beading holds its shape with thread and the form follows the path of the string, and a beaded piece can usually be unstrung and rebuilt. Fuse beads hold their shape through heat and the form follows the arrangement on the board, which makes the fusion step effectively permanent. Tools, materials, and finished strength all differ.

The second is believing a longer ironing pass makes a stronger piece. Over-ironing flattens the holes and turns the beads into a slab, destroying the pixel look and thinning the edges until they curl. The goal is to fuse neighbouring beads at their surfaces while keeping the holes open, not to melt the whole thing into sheet plastic.

The third is assuming color codes can be converted between brands. Codes are each manufacturer's own numbering, so there is no conversion between P01 and H01. The correct approach when switching brands is to re-match by color, which is covered in the brand color guide, and to do that matching against the actual beads rather than printed swatch charts, whose reproduction varies between screens and printers.