The Recipe Graveyard: We Priced 4,763 Recipes and 434 Pay

Ingredient lists are everywhere. A number at the end of one is not. Every recipe has two sides, what the reagents cost and what the finished item sells for, and almost nobody puts those two sides on the same page.

So we did it across the whole book. Cost to craft is the sum of each reagent at its cheapest current auction price. Crafted value is the live auction price of the finished item. Profit is the second minus the first minus the 5% Blizzard keeps on a successful sale. Every figure is recomputed from the latest snapshot, never written down.

Here is what came back.

The census

Measured 31 August 2026	Recipes	Share of the book
In the priced recipe book	4,763	100%
Output has no auction price at all	3,041	63.8%
Output priced, at least one reagent not	112	2.4%
Priced on both sides	1,610	33.8%
Of those, showing a profit	870	18.3%
Profit and a real market behind it	772	16.2%
And clearing ten gold a craft	434	9.1%

Read the top two rows first, because they are the finding almost nobody mentions. Two thirds of the recipes we hold cannot be priced at all. The finished item has no auction listing, so there is no honest number to give you. We leave those blank rather than guess, which is why the recipe book shows a dash on so many rows instead of a confident figure.

The reason is structural. The region-wide auction house carries stackable goods: materials, consumables, reagents, glyphs. Gear does not trade there. It trades realm by realm, so there is no single regional price for a crafted chest piece the way there is for a flask. That single fact bends the entire audit, and it hits the gear professions hardest. Blacksmithing has 600 recipes in the book and only 48 of them can be priced on both sides. Leatherworking manages 47 out of 599. Alchemy, whose output is almost entirely consumables, manages 419 out of 517.

So the honest scope of everything below is this: it is a census of tradeable, stackable crafting. It says nothing about whether forging a weapon for a guild mate was worth your afternoon.

The graveyard

Of the 1,610 recipes we could price on both sides, 740 lose money. Not marginally. In 701 of them the reagents alone ask more than the finished item does, before Blizzard's cut is even applied.

Four hundred of those 1,610 have an output priced under a single gold while their reagents cost real money.

The clearest case in the whole book is Transmute: Primal Fire to Earth. It consumes one Primal Fire and returns one Primal Earth. That is the entire recipe. Open the two prices side by side:

One item goes in, one comes out, and the arithmetic has been pointing the same direction for a long time. Nothing about that recipe is broken. It is a Burning Crusade transmute whose input is still farmed and traded and whose output stopped being wanted years ago.

Wildvine Potion is the same shape from the other direction. It needs Wildvine, a classic-era herb almost nobody gathers now, plus Purple Lotus and a Crystal Vial. The herb alone is the expensive part. The potion it makes, Wildvine Potion, is a consumable with a required level of fifteen, and nothing in the current game asks for it.

Flask of the Titans is the subtle one, and the most instructive. Its output has genuine collector demand, so it carries a real price. It also needs Black Lotus, which is one of the scarcest herbs in the game. The finished flask sells for a lot. The lotus in it costs more.

Why a recipe dies

Every dead recipe in the list has the same anatomy: the inputs live in a market that still moves, and the output lives in a market that stopped.

That asymmetry is permanent by design. Old reagents keep a price because they are gated behind farming that few people do, and because transmutes, collectors and legacy achievement runs keep pulling small amounts out of supply. Old outputs lose their price because nothing consumes them. Demand for a level fifteen consumable does not come back.

The pattern is not limited to classic content. Vantus Rune: Vault of the Incarnates is far more recent, and it requires level 70 to use. It takes two Burnished Ink and one Chilled Rune, both of which are still being made and traded because later recipes want them. The rune itself buffs a raid nobody is progressing any more. Same anatomy, shorter timeline.

When you see a recipe recommended, that is the question to ask about it. Not "is this profitable" but "does anything still consume the output, and is anything else still bidding for the inputs."

The family that quietly pays

The audit turned up one clear winner, and it is not the one the lists talk about.

Of 102 glyph recipes we can price on both sides, 95 are profitable and all 95 have a real market behind them. That is the highest hit rate of any family in the book by a wide margin, and it is not close. Transmutes as a family manage 37 profitable out of 77 priced. Flasks manage 13 out of 37. Potions manage 16 out of 49.

Glyphs win on cost structure. A cosmetic glyph is pigment plus parchment, both of which are byproducts of milling herbs people are already gathering, and the finished glyph is a permanent cosmetic that a player buys once and never consumes. Supply is cheap and demand never fully clears.

Glyph of Fel-Touched Shards is representative. One Light Parchment, one Sallow Pigment, one Roseate Pigment:

Compare the three reagent prices to that one and the margin is obvious without any arithmetic from us. The whole Inscription book is worth a slow read for the same reason.

The method, so you can run it yourself

Four numbers decide whether any craft is worth doing, and only two of them are the ones people quote.

1. The output ask. The cheapest current listing for the finished item. Not what it sold for. There is no public sale feed in this game, so every price on every site is an ask, including ours.

2. The reagent asks. Sum each reagent at its cheapest listing. Where a slot accepts several materials, price the cheapest one that is actually available, because that is what you would really buy.

3. The cut. Blizzard keeps 5% of a successful sale. On a thin margin that is the whole margin.

4. Depth. This is the one that gets skipped and it is the one that matters most. A margin computed against a single listing is not a market, it is one person's asking price. Our own gate is three or more competing auctions and at least ten units listed region wide. Applying it removes 98 recipes that otherwise looked profitable, and it removes the largest headline numbers first.

Depth is also why "profitable" is a weaker word than it sounds. Of the 772 recipes that pass both profit and depth, 338 net under ten gold per craft and 134 net under one gold. Positive is not the same as worth your time. That is how the census gets from 870 down to 434.

What this does not measure

Being explicit about the edges is the difference between an audit and a claim.

There is no margin history. Cost to craft is computed live, every time the page loads. We do not store a rolling record of what any recipe's margin was last month, so nothing here can tell you a recipe has never paid. It can only tell you what it does at the current snapshot. When this page says a recipe loses money, read that as now, not always.

Prices are asks, not sales. Every figure on this site, and on every site, is derived from what sellers are asking. Nobody has the sale data.

Crafting skill is not modelled. Multicraft, resourcefulness and concentration all change the real economics of a modern craft, sometimes by a lot. The census prices one craft at listed material cost with no procs. If you have invested in a specialisation, your true cost is lower than the number here. Crafting with concentration covers that side.

The book is not every recipe Blizzard has shipped. The recipe index returns 11,040 recipes, 10,085 of which declare reagents, producing 5,868 distinct crafted items. Our book prices 4,763 and leans toward older content in several professions. Coverage is being extended.

Checking any recipe before you craft it

Open its page in the recipe book and read in this order. Depth on the output first, because if nobody is buying, nothing else matters. Then the output ask. Then the reagent lines, looking specifically for one line that dominates the total, which is almost always the tell. A recipe with one reagent carrying 90% of the cost is a recipe about that reagent, not about the craft.

The same discipline works on the input side through the materials index. If you gather your own reagents, the material still has a cost, because you could have sold it instead. Crafting with material you farmed is not free, it is a decision to convert one sellable thing into another. How to read auction house price data works through the four numbers in general form, and which ores are worth your time applies the same reasoning on the gathering side.

FAQ

Is crafting profitable in WoW?

For a minority of recipes, yes. Of 4,763 recipes we can price, 772 show a profit with a real market behind the output, and 434 of those clear ten gold per craft. That is roughly nine percent of the book at the snapshot this page was written against.

Why do so many recipes have no profit figure?

Because the finished item has no auction price. Gear trades realm by realm rather than region wide, so 3,041 recipes in the book, mostly from the gear professions, have no regional output price to compute against. We leave those unknown rather than guess.

Which profession has the most profitable crafts?

By count of priced recipes that clear both the profit and depth gates, Jewelcrafting and Inscription lead, followed by Alchemy. Part of that lead is real and part of it is coverage: those professions make stackable goods that the region-wide auction house actually carries, so far more of their recipes can be measured in the first place.

Are glyphs really the best crafting family?

They have the highest hit rate in the book: 95 of 102 priced glyph recipes are profitable and all 95 have real market depth. Individual profits are moderate rather than spectacular, and the market is deep but slow, so treat it as steady rather than fast.

Has this recipe ever been profitable?

We cannot answer that, and neither can anyone else. There is no stored margin history, only a live computation against the current snapshot. Anyone telling you a recipe has always or never paid is describing a memory, not a measurement.