Photo2PatternFree perler bead pattern maker
Photo to perler bead patterns, in codes you can order
Drop a picture, say which beads are in your drawer, and this returns the pattern in that brand’s own catalog numbers — Perler, Hama, Artkal, COCO or MARD, 1,406 colors between the seven lists — with the exact bead count beside every code. It costs nothing and needs no account, and the file is decoded in this tab. The chart arrives split into pegboards rather than paper sheets, with a printable mark in every peg, because the thing you are about to do is clip boards together and count beads into piles.
- 100% free
- No signup
- 7 bead brands
- 1,406 color codes
- Counts per code
Press Ctrl+V anywhere on the page to drop in a sprite or photo already on your clipboard — no need to click the field first.
16 to 232, and the depth follows your picture’s own shape. A single face wants 40 to 60; below about 30 a face becomes a symbol of a face, which is why game sprites read so well at 24 and holiday photographs do not.
Every code in the key is Perler’s own, and one chart never mixes two catalogs. One bead is 5 mm across in the midi class.
Square boards run 29 pegs a side in 5 mm and 57 in 2.6 mm, which comes to 145 mm and 148 mm — the same board, twice the pegs. Change it if yours is a different one.
Beads. Any color needed fewer times than this is removed and its pegs go to the closest surviving shade. Set it to 1 to keep everything. Survivors only ever gain beads, so one pass is enough.
Every cell takes its own nearest color. Flat areas stay flat and the color count stays low — the right choice for anything worked by hand.
How to get a bead pattern you can actually shop for
Only the last decision here saves money — the first two settle how the thing looks.
Drop the picture and name the beads in your drawer
Seven catalogs are offered and the chart is built against exactly one of them, because P17 and H17 are different plastics from different companies and a key that mixed them would be unorderable. Choose the brand you actually own rather than the biggest list: a 291-color catalog produces a smoother face and a shopping trip, while the 117 Perler codes produce a chart you can finish from what is already in the tub.
Set the width in beads and let it land on whole boards
Depth follows from the shape of your picture, so the only number you choose is how many beads run across the top. Leave the board rounding on and the width snaps to a multiple of your board — 29 pegs for a 5 mm square board, 57 for a 2.6 mm one — because a piece that ends eleven pegs into a fourth board still needs that fourth board clipped onto the table.
Cut the one-bead colors before you order anything
Raise the drop threshold until the key stops listing colors you would buy a whole bag of for six beads. Each removal re-matches those pegs onto the nearest shade that survived, so the picture closes over the gap instead of leaving holes, and the surviving counts only ever grow. Then take the CSV, or print one sheet per board and work from paper.
Technical specifications
| Bead catalogs | 7, held separately and never mixed in one key: Perler 117, Hama 89, Artkal 176, Artkal Mini 221, COCO 291, MARD 221 and MARD 291 — 1,406 codes in all |
|---|---|
| Bead sizes | 5 mm midi for Perler, Hama, Artkal and both MARD lists; 2.6 mm mini for Artkal Mini and COCO. Finished width is beads across times that diameter |
| Grid | 16 to 232 beads across — 232 being eight 29-peg boards — with the depth taken from the picture's own proportions |
| Pegboard | 29 pegs a side for a 5 mm square board (841 pegs) and 57 for a 2.6 mm one (3,249 pegs), which measure 145 mm and 148 mm — one board size, twice the pegs. Editable if yours is not |
| Rare-color cut | Removes any color needed fewer than N times and re-matches its pegs onto the nearest survivor. Survivors can only gain beads, so a single pass is provably enough |
| Transparency | Cells averaging over 50% transparent are left as bare pegs, excluded from the bead counts and from the shares rather than filled with white |
| Printing | One sheet per board with no repeated row, since boards butt rather than overlap; peg numbers stay in whole-piece coordinates so board four reads 88, not 1 |
| Exports | PNG of the whole piece or of one board, and the key as CSV with the mark, the brand code, the color name, the hex and the bead count |
Frequently asked questions
Which brand should I pick if my tub is a mix of everything?
Pick the brand whose bags you can still buy more of, not the one that happens to be on top. A chart is a shopping list first and a picture second, and its value is that every line on it names something you can order by code — which stops being true the moment two catalogs are mixed, because nothing on the shelf is labeled with both. If your tub is genuinely unsorted, run the chart against the 117 Perler codes: it is the smallest catalog here, so the shades it picks are the ones most likely to have a match in a jumbled drawer.
Why does the same photo need 41 colors at one brand and 22 at another?
Because a bigger catalog has somewhere finer to send each peg. Matching is nearest-color in Oklab, so a 291-shade list splits a cheek into four tones where a 89-shade list rounds all four to the same one. Neither answer is wrong; the large catalog gives a smoother face and forty-one separate things to buy, and the small one gives visible banding and a piece you can finish this afternoon. The drop threshold exists to move you back down that scale after you have seen what the full list costs.
How big will the finished piece actually be?
Beads across times the bead diameter, and nothing else — 58 beads of 5 mm is 290 mm, whatever the picture was. The pegs set the spacing before any heat is involved, so the arithmetic holds: a 2.6 mm mini bead makes the same 58-bead chart 151 mm instead. Light fusing flattens the tops rather than moving the beads, so a well-ironed piece measures within a millimeter or two of the figure shown; a piece held under the iron until the holes close will curl and pull its edges in by more.
Can a pattern be wider than one pegboard?
Yes, and the chart is split by board rather than by paper for exactly that reason. Square boards clip together, so peg 29 and peg 30 are on different boards with nothing between them, and the printed sheets carry no overlapping row — an overlap would be a lie about where the join is. Work each board complete, iron it, then fuse the seams with the boards butted; that order is more forgiving than trying to lay the whole picture out first.
Why are some pegs left empty instead of being filled with white?
Because a cell that was more than half transparent has no color to round to, and guessing white would put beads on your bill that the picture never asked for. Anything below that threshold is left bare, excluded from every count and shown as a pale square. This is the argument for feeding the tool a PNG cut out to its subject: a photo on an opaque background gets charted as a rectangle, background included, and the background is usually most of the beads.
Should I turn error spreading on?
Not for anything you are placing by hand. Diffusion buys a closer average color by scattering single beads of a fourth and fifth shade through a flat area, which is invisible at arm's length and costs you two more bags and a lot of squinting at the chart. It is off by default here for that reason. If you want texture, take the ordered pass instead: its crosshatch repeats, so it reads as a weave you chose rather than as beads that got away from you.
What are the marks in the squares for?
They are what makes the chart usable on a grayscale printer and by anyone whose eyes do not separate two adjacent shades, which is a twelfth of the men who will look at it. Every color in the key gets one of 70 marks, handed out so the colors you place most often get the simplest shapes to scan for. Two mid-grays that look identical on a printout are still an unmistakable circle and an unmistakable triangle, and that is the whole distance between a chart somebody can follow and a pretty grid.
About fuse beads as a fixed palette
Fuse beads are the most honest material on this site. Embroidery floss has no published screen color at all, and paint is mixed rather than picked, but a bead is injection molded in a cataloged shade with a code printed on the bag — so the palette a chart snaps to is a real finite set, and every line of the key is something a shop can hand you. That is also the trap. Because the list is exact, people treat the number of colors as free, and it is not: forty-one shades in a key means forty-one bags, forty-one piles on the table and forty-one chances to reach into the wrong one. The single most useful control on this page is the one that removes colors, not the one that adds them.
The other thing most bead generators get wrong is geometry. They resample a picture into a square of arbitrary width and hand it back, as if the grid were a screen resolution. It is not — it is a set of physical boards that clip together, 29 pegs a side in the 5 mm class and 57 in the 2.6 mm class. Those come to 145 mm and 148 mm respectively, which is the point: it is one board size, and the smaller bead simply buys twice the pegs inside the same square of table. A pattern 100 beads wide is not a 100-pixel image; it is three full boards and a thirteen-peg strip, and the strip is the part that ruins an afternoon. Rounding the width to whole boards moves it to the nearer of 87 and 116 — three boards or four — and either one is a piece with no strip in it. If you want the finished piece costed before you commit — bags per color, waste, the difference between the naive division and the real one — that is the bead count calculator, which starts from a chart rather than making one.
Two neighboring pages run the same match against a different fixed set, and comparing them is the fastest way to see what a palette does to a face. Minecraft’s map colors reach 61 shades on a flat build, so the Minecraft version bands more visibly and rewards a bigger grid, while a brick portrait is worked against fifty colors still in production and has to survive an 8 mm stud pitch, so it needs far more width to read at all. If your beads are the pony kind strung on elastic rather than the fuse kind ironed on a board, the grid is offset instead of square and the kandi pattern maker is the page that accounts for it. And if you arrived here with pixel art that is already the right shape but the wrong size, resize it on whole-number multiples before you chart it, so no original pixel is split across two beads.
Product and company names shown here are the trademarks of their respective owners. Photo2Pattern is an independent tool and is not affiliated with, sponsored by or endorsed by any of them.
The photo stays on the peg board side of the screen
Your picture is opened by JavaScript running in the tab you are reading this in. It is never uploaded, never sent to a server and never stored anywhere — there is no upload step to undo, because there is no upload. Close the tab and nothing of it remains.
The brand you choose, the board size you type and the chart itself live in this tab and nowhere else — nothing is remembered between visits, so take the CSV or the PNG if you want the pattern to outlast the session.