Why Grams to Cups Converters Disagree
This site says a cup of flour is 120 grams. Several well-known converters say 125 grams. Both figures are defensible. Neither site usually explains itself, so here is the comparison laid out.
Values below were read from each site in September 2026 and are attributed to them. Where a site does not publish a figure, the cell says so.
| Ingredient | GramstoCup.co | gramstocup.com | gramstocups.io | gramstocupsconverter.cc | TheCalculatorSite | What explains the gap |
|---|---|---|---|---|---|---|
| All-purpose flour | 120 g | 120 g | 125 g | 125 g | 125 g | King Arthur's tested spoon-and-level figure against the more common US value. |
| Granulated sugar | 198 g | 200 g | not published | not published | not published | 200 g is a rounding of the same measurement; the real gap is about 1%. |
| Brown sugar, packed | 213 g | 220 g | 220 g | not published | not published | Packing pressure varies more between bakers than these figures do. |
| Butter | 227 g | 227 g | not published | not published | not published | Universal agreement. Butter is sold by weight, so there is nothing to disagree about. |
| Cake flour | 114 g | not published | not published | 114 g | not published | Agreement here, even where the all-purpose figures differ. |
| Almond flour | 96 g | not published | not published | 96 g | not published | Agreement. |
| Fresh whole milk | 244 g | 237 g | measured density | not published | not published | 227 g and 237 g both come from treating milk as a volume of water. |
Read from the named sites in September 2026. Figures published by other sites are quoted for comparison and are not part of this site's dataset.
120 grams is King Arthur Baking's tested figure. King Arthur is the largest flour miller in the United States and measures with a stated method: fluff the flour, spoon it into the cup, level the top. That is the value this site uses.
125 grams is the more common figure elsewhere. It sits closer to what a cup weighs when filled a little more firmly than spoon-and-level produces.
The gap is 5 grams, about 4%. Now set that against the variable nobody argues about: the same cup of flour holds roughly 106 grams sifted, 120 grams spooned and levelled, and up to 150 grams scooped straight from the bag. How you fill the cup moves the answer roughly eight times further than which reference chart you picked.
That is the useful conclusion. Arguing about 120 against 125 is arguing about the small term while ignoring the large one.
Most charts list a cup of milk at 227 grams, and some at 237 grams. Both come from treating milk as a volume of water: 227 grams is 8 fluid ounces converted as if fluid ounces were ounces of weight, and 237 grams is the weight of a cup of water.
Milk is not water. Fresh whole milk has a measured density of about 1.03 g/mL, which gives 244 grams for a US cup. A chart that lists milk lighter than water has an error you can see without measuring anything.
This is not a lone position. gramstocups.io reaches the same conclusion in its own source notes, observing that King Arthur's liquid rows treat 8 fluid ounces as 8 ounces of weight, and using measured density for liquids instead.
Four questions. A converter that cannot answer them is asking you to take its arithmetic on faith:
| Question | What a good answer looks like |
|---|---|
| Which cup? | A stated volume. 236.588 mL for a US customary cup. |
| What grams-per-cup value? | A number you can see, not just the final cups figure. |
| Which product state? | Packed or loose, sifted or not, cooked or uncooked. |
| What source? | A named, checkable reference, with the date it was checked. |
This site publishes all four for every ingredient on the density data page, along with a downloadable dataset and a methodology document. That is the standard we would want to be held to.