How it works
Each of the eleven blends is stored as a list of ratio parts rather than fixed grams. The calculator multiplies any spice tagged as heat by your chosen multiplier, adds up the adjusted parts, then gives every spice a share of your target weight equal to its part divided by that sum. The consequence is that the batch always weighs exactly what you asked for. The approximate teaspoon column simply divides each weight by 2.5 grams per teaspoon.
The blend itself sets the character: curry powder, garam masala, taco, cajun, Italian herbs, pumpkin pie, ras el hanout, Chinese five-spice, herbes de Provence, jerk and za'atar. Target total weight, 50 g by default, just scales the whole thing. Heat level, offering half, standard, 1.5× and 2.5×, only touches spices marked heat, which means cayenne and the chili powder in taco seasoning. Salt: on or off adds or removes the salt line in taco, cajun, jerk and za'atar.
Because the total is pinned to your target, raising the heat multiplier dilutes every other spice rather than adding chilli on top, and switching salt off makes the remaining blend proportionally stronger per gram. The teaspoon column is the weakest number on the page: it applies one density of 2.5 g per teaspoon to everything from dried lavender to salt, which are nowhere near equal, so trust the grams. These are also one house version of each blend, not a canonical recipe.
Frequently asked questions
What are the ratios in a basic curry powder? +
Four parts ground coriander, three cumin, two turmeric, then one each of fenugreek, mustard powder, ground ginger and black pepper, plus half a part cayenne. That is a mild British-style curry powder, so raise the heat multiplier if you want it fiercer.
Does turning the heat level up add more cayenne? +
It raises cayenne's share, not the batch weight. At 2.5× the cayenne in taco seasoning goes from 0.5 parts to 1.25 and the chili powder from 4 to 10, so every other spice shrinks proportionally to keep the total at your target.
How do I make taco seasoning without salt? +
Press the Salt button to switch it off. The two parts of salt drop out of the taco blend and the remaining eight spices are recalculated to fill the same total weight. Cajun, jerk and za'atar are the other blends containing salt.
Is Chinese five-spice really equal parts? +
In this tool, yes: star anise, cloves, cassia cinnamon, Sichuan peppercorn and fennel seed each get one part. A 50 g batch is therefore 10 g of each. Many commercial versions lean much harder on star anise and cassia.
How many teaspoons is 10 grams of a blend? +
Roughly four, using the 2.5 grams per teaspoon this calculator assumes for ground spice. Treat it as a rough guide only: fine dense powders such as salt weigh far more per spoon, and dried leaf herbs weigh considerably less.
Ratios travel; quantities do not
A spice blend is defined by its proportions, so scaling one is genuinely linear in a way that scaling a whole recipe is not. That is what makes blends worth batching: mix once at any size and the ratio is preserved.
What does not scale is potency. Ground spices lose volatile aromatics steadily once ground — most are noticeably flatter after six months and largely spent within a year — so a blend mixed from old jars needs more of everything to reach the same intensity, and the ratio that was right for fresh spice may not be right for tired spice.
Whole against ground, and toasting
Whole spices keep their aromatics far longer than ground, which is why grinding to order gives a blend a shelf life measured from the day you make it rather than from the day the jars were filled.
Volume conversions between whole and ground are unreliable — grinding changes packing density unpredictably, and a tablespoon of whole cumin does not produce a tablespoon of ground. Weigh whole spice before grinding if a ratio matters.
Toasting before grinding changes flavour substantially and weight slightly, as moisture is driven off. If you toast, do it consistently across batches, or the same recipe will taste different each time for reasons that look like measurement error.
