I worked this out at eleven at night in June 2019, a few hours after a naturopath told me I was allergic to eggs.
I had been using konjac root powder for a while already, in place of xanthan gum in gluten-free baking. Lying awake that night mourning eggs, it occurred to me that if konjac gels well enough to replace a gum, it might gel well enough to replace an egg. I got up, went to the kitchen, and it worked on the first try, which almost never happens.
Seven years on, this is still the cookie base I come back to. Every other cookie on this site is a variation of it. The formula below is the one I have settled on after making it several hundred times, and the substitution tables underneath cover every version I have published separately over the years.
Why konjac and not flax or chia
Glucomannan, the fiber in konjac root, holds a great deal of water in a gel. That gel does roughly what an egg white does in a batter, which is hold everything together while it sets.
The reason it beats the usual replacements is what it does not do. Flax carries a faint fishiness that reads clearly in a cookie even where it disappears in bread. Chia speckles the dough and darkens it. Banana and applesauce both bring their own flavor and their own sugar, and what comes out is banana bread in cookie form. Konjac brings nothing. It is a common ingredient across a lot of Asian cooking and has been for a very long time, and it behaves the same way every time.
Eggs work too, and here is when to use them
Three eggs in place of the three konjac eggs, and nothing else in the recipe changes. What you get is a lighter, fluffier cookie, a little more protein, and a texture closer to what most people expect. The konjac version is denser and chewier, which is the one I prefer, and it is the one to make if eggs are off the table for you.
The other thing worth saying plainly is the cost. A teaspoon and a half of konjac powder runs to a few cents. Three eggs, at what eggs cost now, do not. If the constraint is the grocery budget rather than an allergy, the konjac version is the cheaper cookie by a wide margin and you are not giving anything up to get there.
How much water your konjac egg needs
One konjac egg is half a teaspoon of powder in a quarter cup of water, as a large egg is about a quarter cup. Three eggs is a teaspoon and a half in three quarters of a cup.
Tahini is the reason that is a range rather than a number. Some jars pour and some need a spoon and real effort, and the same weight of each behaves differently in a dough. Start with half a cup of water. If your tahini is loose, stop there. If it is stiff, add the last quarter cup. This is the one adjustment in the recipe that actually matters and it is why earlier versions of this post carried different numbers.
Which flour, and what changes
The flour is the part I have swapped most, mostly because of what I could digest in a given season. They are not interchangeable by volume and they do not behave the same.
- Coconut flour, 1/3 cup. The default below and the most absorbent of them by a wide margin. Nothing else needs so little.
- Sesame flour, 1/2 to 3/4 cup. What I used through the years I could not tolerate coconut. It doubles down on the tahini and it is the closest in behavior. Start at half a cup and add more if the dough will not hold a scoop.
- Tigernut flour, 3/4 cup. A tuber rather than a nut, and naturally sweet. This one is different enough that it has its own post, as tigernut with a little allulose makes a dough that spreads by itself and needs no pressing.
- Chestnut flour, 3/4 cup. The closest substitute for tigernut. Similar sweetness and similar starch. It browns faster, so pull them a minute or two early, and go lighter on the monk fruit.
- Almond flour, 2/3 cup plus 2 tablespoons coconut flour. Almond alone has nowhere near enough absorbency and you get a puddle. The coconut flour is doing the job the fiber in the other flours does. Chill the dough fifteen minutes before scooping.
Cassava and arrowroot do not work here as the main flour, which is worth saying as they are the two most people reach for in grain-free baking. Both are close to pure starch, and in a dough carrying this much fat they go gummy in the middle and stay pale on the outside. A tablespoon of arrowroot is useful for tightening a dough that has come out too loose, and that is the only job either of them has in this recipe.
Collagen, or maca if you are vegan
Collagen is what gives these the chew I spent years chasing. It dissolves into a dough without going chalky or squeaky the way whey protein does in baking, and it has no flavor of its own. It also means a cookie here carries more protein than most people expect from a cookie, which I mention to my own family and to nobody else.
Collagen is not vegan. After a long stretch of trying alternatives, gelatinized maca is the best chewy substitute I have found, at three tablespoons in place of the six scoops. It also brings a malted flavor that is genuinely good rather than merely acceptable, and two of my three kids prefer it. Red maca and black maca both work. A blend of sacha inchi, pea protein and maca works too if you want the protein higher.
Maca absorbs a lot of liquid, so add it a teaspoon at a time past the first three tablespoons.
Monk fruit is not a standard strength
Pure monk fruit extract sold to home bakers runs anywhere from seven to fifty percent mogroside V, which is the compound doing the sweetening. That is a sevenfold spread between the weakest jar and the strongest, and almost nobody prints the number on the label. A teaspoon of one brand is not a teaspoon of another.
This is why the amount in this recipe has moved over the years. My first brand was concentrated enough that a level teaspoon was right. It was discontinued for a stretch, and the organic one I replaced it with needed half again as much. Where I have landed across three brands:
- A concentrated extract — 1 teaspoon.
- A milder organic one — 1.5 to 2 teaspoons.
- Two teaspoons for other people. I like things less sweet than most, and two is the number that works when these go to a potluck or get handed to somebody else’s kids.
Taste the dough. If you are using the konjac egg there is no raw egg in the bowl, so you can taste it as much as you like and adjust before anything goes in the oven. That is the only reliable way to calibrate a new jar, and it takes ten seconds.
If you want them sweeter, do not add more monk fruit
Two teaspoons is about the ceiling. Past that most brands develop an off note, a faintly cooling bitterness at the back of the mouth that gets worse the more you add, and no amount of it will make the cookie taste the way a sweeter cookie tastes. What is missing at that point is bulk, not intensity. Add a little of an actual sweetener alongside the monk fruit instead, and keep the monk fruit where it was.
- Maple sugar, 2 to 4 tablespoons. Dry, so nothing else in the recipe changes. It behaves the most like sugar and browns like sugar. This is what I use when these are going to a potluck or to somebody else’s house, where people are not as sugar-sensitive as I am.
- Allulose, 2 to 4 tablespoons. Powder it in a blender first if yours is granulated. It keeps a cookie soft and it browns considerably faster than sugar does, so pull them a minute or two early or drop the oven fifteen degrees.
- Maple syrup or honey, 2 to 3 tablespoons. Both are liquid, so take the same amount back out of the konjac water. Both brown quickly, so watch the last few minutes. Honey is not vegan, and honey that has been to 325°F is not raw in any sense that matters, so use the ordinary jar for this rather than the good one.
Any of these adds carbohydrate. That is the trade and it is a fair one, as a cookie somebody actually enjoys is worth more than a cookie that is technically perfect and gets left on the plate.
The chips I actually use
Three brands, sweetened three different ways, and the difference matters more than it sounds. The chips are usually the largest single source of carbohydrate in this cookie, so whichever bag you reach for is doing more to the final numbers than anything else you change.
- Just Date. Dark chips sweetened with dates, and the easiest of the three to find. Dates are sugar, so these are the ones that put real sugar into the cookie. In a whole-food kitchen that is often the right trade and I make it happily. If you are watching blood sugar closely, this is the bag to know about.
- Goalz. Allulose sweetened, which is largely not absorbed, so these barely move the carb count. The dark and the semi-sweet are both excellent and the milk chocolate is genuinely good, which is unusual in this category. They also make white chips and seasonal butterscotch and peppermint, and those are the ones the low-temperature bake below exists for.
- ChocZero. Monk fruit sweetened, so nothing on the sugar side at all.
I use the dark in most things. My kids prefer the semi-sweet or the milk in a cookie, and I stopped arguing about that a few years ago.
Your chips set the oven temperature, not the dough
This is the thing that took me longest to work out and it applies to every cookie on this site. The batter bakes fine at 325°F. Some low-sugar baking chips do not, and they collapse into a greasy mess well before the cookie is done.
- Dark and semi-sweet chips — 325°F for 10 to 13 minutes.
- White chocolate chips — 270°F for 10 to 12 minutes.
- Butterscotch and caramel chips — 225°F for 10 minutes. These are the fragile ones.
The low-temperature versions do not bake hard, so give them a full ten minutes on the tray before you move them or they fall apart.
Tahini Chocolate Chip Cookies
The base recipe everything else here is built on. Grain-free, no refined sugar, and egg-free with a konjac egg that brings no flavor of its own.
Conversions are approximate. Whole foods vary.
Nutrition — tahini and collagen together carry real protein and real minerals, which is not usual for a cookie and is most of why this one exists.
Blood sugar — no sugar, and the fat and protein alongside the monk fruit mean these behave nothing like a bakery cookie.
Let the konjac gel a full two minutes. Glucomannan needs the time to hydrate. Rushed, it stays grainy and the dough will not hold.
Taste the dough before it goes in. Monk fruit strength varies enormously between brands, and the konjac version has no raw egg in it, so tasting is both safe and the only way to get the sweetness right for a new jar.
These do not spread. Scoop them, then flatten each one with your palm. A little coconut oil on your hands stops the sticking.
Check your chips before you set the oven. Butterscotch and caramel need 225°F, white chocolate needs 270°F, and everything else takes 325°F.
Ingredients
The binder, choose one
- 1.5 tsp konjac root powder, in 0.5 to 0.75 cup water
- 3 pastured eggs
The chew, choose one
- 6 scoops collagen peptides, about 0.75 cup
- 3 tbsp gelatinized maca powder, for the vegan version
Everything else
- 0.75 cup tahini, well stirred
- 0.33 cup coconut flour, or see the flour table above
- 1.5 tsp monk fruit extract powder, 1 to 2 depending on brand
- 1 tsp vanilla extract
- 0.75 tsp baking soda
- 0.75 tsp pink salt
- 0.25 tsp Ceylon cinnamon, optional
- 0.5 cup dark chocolate chips, see the brands above
Method
- Heat the oven to the temperature your chips need, and line a baking sheet with parchment.
- Put half a cup of water in a medium bowl and sprinkle the konjac powder over the surface. Whisk, and leave it two full minutes to gel. If your tahini is stiff, use the full three quarters of a cup. If you are using eggs, whisk them and carry on.
- Add the tahini and whisk until it is smooth.
- Whisk in the monk fruit, baking soda, salt, vanilla and cinnamon until the mixture is creamy.
- Fold in the coconut flour and the collagen or maca with a spoon.
- Fold the chocolate chips through. Dried cranberries or chopped dates work here too.
- Scoop onto the parchment. Oil your palms lightly and press each one flat, as this dough will not spread on its own.
- Bake 10 to 13 minutes. Ten for a small scoop, twelve or thirteen for a medium one.
- Cool ten minutes on the tray before moving them.
Minerals
Calcium — notable for a cookie. Sesame, and it is most of why tahini is worth the price.
Copper — high. Tahini, with a little more from the chocolate.
Magnesium — moderate. Tahini and chocolate together.
Zinc — moderate. Tahini.
Protein — high for a cookie. Collagen and tahini, and there is no flour diluting it.
Qualitative, read off the ingredients rather than computed, as the flour and the chips both move considerably between versions.
Notes and swaps
A cookie scoop is worth owning for these. The dough is wetter than a conventional one and does not roll well by hand.
A food processor works: everything but the flour and the chips goes in, blend smooth, then pulse the flour in and fold the chips by hand.
Monk fruit strength varies sevenfold between brands. Taste the dough and adjust before it goes near the oven.
Three eggs in place of the konjac gives a fluffier cookie and a little more protein. Konjac gives the chewier one, and costs a fraction of what three eggs cost.
For a sweeter cookie, add 2 to 4 tbsp maple sugar or allulose rather than more monk fruit.
Cranberries or chopped dates in place of the chips make a different and more Middle Eastern cookie. Cardamom belongs in that version.
Black tahini in place of the regular gives you a near-black cookie. Caramel chips suit it particularly well.
They keep four or five days in a tin and freeze well scooped and raw on a tray.
Brenna May Wellness
The other cookies built on this
- Dark Chocolate Espresso Tahini Cookies, which swaps the flour for raw cacao and takes mint or chaga variations.
- Tigernut Tahini Chocolate Chip Cookies, the version that spreads on its own and needs no pressing.
Where I get these
The collagen, the tahini and the konjac powder are all in my dispensary. Konjac is worth buying somewhere you trust in particular, as it is sold in a range of grinds and only the fine one gels properly.
The rest of the pantry is there too, along with the household cleaning and skincare I actually use. Browse the whole dispensary.
Some of these links earn me a small commission. It does not change what you pay, and I only list things I actually use.