Most gelato makers choose their hazelnut paste once every few years, then live with the decision. It’s worth understanding what you’re looking at.
This guide starts from an honest premise: we make it ourselves, so it isn’t an impartial comparison between brands. What we can do is explain the parameters that genuinely matter, tell you what ours are, and leave the comparison to you.
Why it costs more than an industrial paste
It’s the first question we get asked, so let’s answer it straight away: yes, it costs more. It’s worth understanding what that difference is made of.
Comparing the price per kilo of an artisan paste with an industrial one is like comparing artisan gelato with industrial gelato by the kilo. The numbers compare; the products don’t.
The calibre of the nut
We use only hazelnuts of 11/13 and 13/15 mm calibre, never below 11. This is where much of the price gap comes from: undersized and broken hazelnuts cost far less, and anyone using them can sell for less.
Why calibre changes the flavour, not just the cost
A whole hazelnut is almost spherical. That means the centre is equidistant from the inner surface: heat travels through it evenly, and the roast reaches everywhere in the same way.
A broken hazelnut doesn’t have that geometry. Its inner fibres are exposed, and roasting it is like cooking them directly. On top of that, with more surface in contact with air it oxidises sooner: part of the aroma is gone before the nut even reaches the roaster. And pieces of different sizes roast unevenly — some scorch, some stay behind.
The result is a raw material that has lost aromatic intensity. And a lost aroma has to be compensated for: that’s why many industrial pastes contain added flavourings. It isn’t a flavour choice, it’s a remedy.
One roasting curve per calibre
We don’t roast everything together. Each calibre has its own dedicated roasting curve, and for each one there are three grades: light, medium and dark. Putting different calibres in the same roaster means burning part of the batch and leaving part of it underdone.
Our pastes are a balanced blend of calibres, adjusted to what each harvest yields. That’s why we don’t declare a single calibre on the label: it would be a made-up number.
First things first: pure paste or semi-finished product?
This distinction changes everything else.
- Pure paste: only roasted, ground hazelnuts. No sugar, no added oil, no flavourings. The ingredient list has one line.
- Semi-finished or sweetened paste: already contains sugars, and often oils or emulsifiers. It costs less per kilo, but you’re also buying sugar you could have dosed yourself.
With pure paste you decide the structure of the recipe: sugars, fats, solids. With a semi-finished product you inherit someone else’s choices. If you have a balanced recipe you care about, pure paste is the only route that doesn’t disturb it.
One commercial detail worth watching: percentages. A “hazelnut” paste at 50% is half something else. 100% means 100%.
Roast level decides the flavour, not the quality
This is the most common misconception: thinking the dark roast is “better” or “more intense” in absolute terms. It isn’t — it’s a different profile, chosen according to the flavour you want.
| Roast | Profile | When to use it |
|---|---|---|
| Light | Fine, delicate aroma, notes of fresh hazelnut | Clean-tasting gelato, pale creams, semifreddi where the hazelnut shouldn’t dominate |
| Medium | Balance between natural sweetness and roasted intensity | The most versatile: if you can only pick one, pick this |
| Dark | Decisive roasted notes, long finish | Bold flavours, pairings with cocoa, where you need persistence |
A practical note: the dark roast holds up better alongside chocolate, which otherwise tends to cover the hazelnut. The light roast, conversely, gets lost next to strong ingredients.
As mentioned above, each of the three grades comes from distinct roasting curves per calibre: it isn’t the same roast pushed further or less far, they are separately built profiles.
Dosage: always relative to your base
For 100% pure paste the range we recommend is 80–100 grams per kilo of mix. But the dosage on its own means nothing: it has to be judged together with your gelato base.
The reference isn’t how many grams you add, it’s the final fat value you want in the finished product. Pure paste is over 60% fat, so every gram you add shifts that figure.
| Base | Dosage | Why |
|---|---|---|
| Standard | 80–100 g/kg | The working range. Above 100 g/kg the fat rises and the structure changes: the mix needs rebalancing |
| Lean | higher | There’s room on the fat side: the hazelnut can go up without unsettling the structure |
| Rich | lower | The paste’s fat adds to what’s already there, so you reach the threshold sooner |
Below 80 g/kg on a standard base the hazelnut is noticeable but not the lead: that makes sense when it’s a supporting ingredient inside a composite flavour, not when it is the flavour.
If you’re working with a sweetened paste the dosages aren’t comparable: that one brings sugars this one doesn’t, so it shifts your sugar balance as well as your fat balance.
Particle size: watch how it is declared
This is the fineness of the grind, measured in microns (µm). It determines whether you perceive grain on the palate — and it is the parameter most easily misread.
Many suppliers state “under 20 microns”. That figure is almost always the refiner manufacturer’s spec: what the machine is capable of, not what is in the tub. A hazelnut paste contains particles of different sizes — some below 20 microns, some above. There is a spread, and a single number doesn’t describe it.
We state below 25 µm because 25 microns is the size of our filter: a threshold nothing coarser gets past. Our pastes are 25 microns and down, and that is verifiable.
A lower number looks better on paper. It is worth asking whoever quotes one: is that the machine’s capability, or the size of the filter? The two are not equivalent, and only the second tells you what is actually in the product you’re buying.
Oil rising to the top isn’t a fault
A 100% hazelnut paste contains no emulsifiers, so the oil naturally separates and rises to the surface. That’s the correct behaviour: it means there’s nothing inside holding it in place artificially.
Just stir before use, working the oil back into the mass. If a 100% paste never separates, it’s worth asking why.
Storage
- Cool and dry, maximum 25 °C, away from light.
- No refrigeration needed: cold hardens the paste and makes it difficult to dose.
- The real enemy is oxidation. Close the tub properly and don’t leave it open.
- Our pure paste has a shelf life of 24 months from production.
Three mistakes we genuinely see
- Comparing price per kilo across different products. By far the most frequent. A pure paste and a flavoured industrial one are not the same product: in the second case you’re also buying sugar, flavourings and — often — reject-grade nuts.
- Reusing the dosages of your previous paste. This is the only genuine return we’ve had: a customer applied the dosages he used with a sweetened paste to our pure one, and the mix fell apart. When you change paste, the balance has to be redone.
- Changing roast without adjusting the recipe. Going from light to dark shifts the flavour markedly: it isn’t a neutral substitution.
Our pastes, briefly
Italian hazelnuts grown and processed by us in Avella, Irpinia. Only 11/13 and 13/15 calibres, dedicated roasting curves per calibre, three grades, refining below 25 µm. Tubs of 5 and 10 kg, plus a 1 kg format for small productions. No minimum order: the minimum is one full pack.
All the technical data — allergens, nutritional values, packaging, delivery times — is on the pure paste technical sheet.
We ship across Europe and, for trade orders, worldwide via groupage freight: we have already shipped to Japan and the United States.