Read this guide in
FROZEN DESSERT GUIDE
Natural Sweetener Blends for Ice Cream and Frozen Desserts
Build a reduced-sugar ice cream sweetener system: allulose for bulk and freezing behavior, monk fruit for sweetness, with evidence, trade-offs and controlled trials.
An allulose–monk fruit blend is one direction for reduced-sugar ice cream: allulose contributes ingredient mass and affects freezing behavior, while a small amount of monk fruit extract can raise sweetness. The complete formula must still balance water, solids, fat, protein, air and the stabilizing system to reach the intended scoopability and melt profile.
This guide is for dairy ice cream, plant-based frozen dessert, gelato, sorbet, frozen yogurt, frozen novelty and private-label buyers comparing monk fruit, stevia, allulose or erythritol-free sweetener systems.
Judge the frozen product at its serving condition
Compare scoopability, texture, melting and flavor after the intended freezing and hardening process. The drawn Blend A/B/C cups, tick marks and dial illustrate checks to plan, not tested formulas, a ranking or a measured freezing point. The scoops and cabinet do not establish a recipe, production capability or shelf life.
Define the frozen dessert format
A dairy ice cream, high-protein pint, plant-based dessert, sorbet, gelato, frozen bar and soft-serve mix each needs a different balance of sweetness, solids, fat, protein, stabilizer and freezing behavior.
Separate sweetness from texture
Monk fruit and stevia can provide sweetness at low use levels, but they do not replace the grams of sugar removed or supply the same bulk contribution at those doses. Wang, Sala and Scholten's model study related hardness to both ice content and serum-phase viscosity. Melting duration was more closely related to ice content, while viscosity had a greater influence on the delay before melting began. This model helps separate the measurements needed; it does not predict a finished commercial ice cream.
Why test allulose with monk fruit in ice cream?
Allulose and monk fruit extract solve different parts of the reformulation problem. Cargill describes allulose as about 70% as sweet as sugar and identifies bulking and freezing-point depression among its functions. This explains why an allulose-based system may need additional sweetness even when it supplies a useful amount of ingredient mass. The nominal sweetness figure is not a complete ice-cream conversion formula.
| Part of the system | Main role to evaluate | What the ingredient alone does not settle |
|---|---|---|
| Allulose bulk ingredient | Contributes mass and dissolved solids; affects freezing behavior and the sweetness profile | The required addition, serving hardness, melting and storage texture of the complete product. |
| Monk fruit extract | Adds sweetness at a comparatively small addition; assess onset and finish | The mass of sugar removed, the complete solids balance or the stabilizing system. |
| Complete ice-cream formula and process | Balances water, fat, protein, other solids, stabilizers/emulsifiers and incorporated air | Performance after the actual freezing, hardening, storage and serving conditions still needs checking. |
Hengliyuan’s allulose-based, monk-fruit-sweetened direction is a candidate to compare with the current product. If V50 is being considered, confirm the extract identity, Mogroside V assay and method, lot and actual addition. An extract grade does not specify its proportion in the ice-cream blend. Judge the sweetness and finish in the hardened product; do not assume a generic monk fruit potency tells you the dose of a particular supplied grade.
A practitioner’s rating is an application opinion
In her Instagram video, @ambergelatoo rates five ingredients for ice cream with no added sugar. Her preferred direction is allulose paired with monk fruit: bulk and texture support from one ingredient, additional sweetness from the other. Her caption explicitly describes the ratings as based on her own experience and says the result depends on how and where the ingredients are used.
| Ingredient | Her rating out of 5 | Her stated application view |
|---|---|---|
| Erythritol | 2 | Cooling sensation and crystallization concerns in ice cream. |
| Sorbitol | 3.5 | Smooth, creamy texture; caution about tolerance at higher additions. |
| Monk fruit extract | 3 | Very sweet at a small dose, but does not replace the missing bulk. |
| Sucralose | 2 | Efficient and economical; needs careful dosing and does not supply bulk. |
| Allulose | 5 | Her preferred bulk-sweetener direction, paired with monk fruit to complete sweetness. |
These stars express one practitioner’s judgement, not laboratory measurements, a universal ingredient ranking or results for Hengliyuan products. They do not select a ratio or establish a finished-pack claim. The technical reasons and the contrasting supplier example below need their own evidence.
A brand example: product-line wording is a separate decision
KELAO’s original explanation distinguishes its no-added-sugar line from its traditional line and explains why no refined sugar is a different description. The brand mentions monk fruit, allulose and erythritol as possible sweetening ingredients; it does not disclose a complete formula for every flavor.
| Public brand example | What it contributes | How to use it in development |
|---|---|---|
| KELAO: two differently described product lines | The brand’s own caption distinguishes a no-added-sugar line from its traditional line and explains that no refined sugar is a different description. | Keep the intended finished-product claim separate from sweetener selection. Obtain the complete formula, ingredient declaration and market-specific review before choosing pack wording. |
Keep the intended finished-product claim separate from sweetener selection. Obtain the complete formula, ingredient declaration and market-specific review before choosing pack wording. Do not reverse-engineer the brand’s recipe, infer an erythritol-free claim from the product-line name or treat the example as proof of a universal ingredient choice. KELAO is not presented as a Hengliyuan customer or endorser.
Lowering the freezing point is not a guarantee against ice crystals
Freezing-point depression changes how much water remains unfrozen at a given temperature. It is one reason a sweetener change can alter scoopability, but a softer product does not by itself prove less ice-crystal growth during storage. Measure the hardened product at the intended serving temperature and check its response to the relevant storage and temperature fluctuations. The distinction also prevents an overly simple ban on erythritol. Jungbunzlauer’s published light-protein-ice-cream work describes crystallization and hardening risks, yet reports that its allulose-only formulation had poor texture and that a combination with erythritol improved texture and scoopability. Its formulations also involved ingredients such as stevia and inulin. This is a supplier application report on particular systems, not an isolated two-ingredient test, a recommended ratio or evidence that every erythritol-containing formula will work. For an erythritol-free brief, keep that restriction and develop the other parts of the formula. For a brief that permits erythritol, its addition and the full system remain a test question. An erythritol-free preference and a universal claim that erythritol cannot be used in ice cream are different statements.
Compare the pairing without hiding the other formula changes
Use a candidate sequence that answers the actual question. Where the project allows it, compare the current reference, an allulose-based candidate and a separately coded candidate with monk fruit added to reach the intended sweetness. The allulose-only candidate is diagnostic; it need not be an acceptable finished product.
| Comparison | What to keep or record | Question answered |
|---|---|---|
| Current reference versus allulose-based candidate | Complete formula by mass, including any water and solids adjustment; the same defined process and serving condition | Which changes in sweetness, hardness, body and melting accompany this reformulation? |
| Allulose-based candidate versus the pairing | Hold the allulose addition and other formula components fixed where feasible; record the exact extract dose and any carrier adjustment | Does the added extract bring sweetness to target, and what happens to the finish? |
| Promising pairing after hardening and storage | Its own formula/version, package, aeration and temperature history; separate taste, hardness, ice sensation and melt results | Does the chosen direction retain the intended eating quality at the relevant checkpoints? |
If the formula is changed to match sweetness, record that change; it is no longer an equal-addition comparison. Do not use a single overall star score to hide a candidate that tastes good but is too soft to hold or too hard to serve. The existing workbook keeps the individual observations and overrun calculation separate.
Use allulose only where the market allows it
Allulose can be useful in permitted markets because it may support body and sugar-like taste while affecting freezing behavior. U.S. label treatment, EU novel-food status and Australia/New Zealand production-route rules are not interchangeable. Confirm the target market, finished food category, ingredient identity and label plan before using an allulose-supported sample direction.
Watch crystallization and aftertaste
Erythritol can bring a cooling effect and crystallization risk in some frozen systems, while stevia and monk fruit can show linger or bitterness if the flavor base is weak. A 2026 Wageningen model found that smaller ice crystals could form a harder, more connected structure in its simplified system, so ice crystal size should not be treated as a single quality score. Check sweetness onset, aftertaste, iciness, scoopability and storage texture together. Source: https://edepot.wur.nl/705184
Start with the frozen dessert and the problem
Tell Hengliyuan the frozen dessert format, final sales market and the sweetness, iciness, hardness or melting issue you want to resolve. Share the known base, ingredient restrictions, process and package; if these are still being developed, describe the intended product and reference. A final formula or order quantity is not required to begin. Before a comparison, identify the material and formula version, fat and protein sources, total solids, stabilizer/emulsifier system, aeration, freezing and hardening route, package and serving condition. After the sample direction is useful, confirm the specification, receiving pack, trial or repeat quantity, requested documents and delivery terms. This is a sweetener ingredient or blend discussion; identify a finished frozen dessert manufacturing request separately so supply fit can be confirmed.
Keep the comparison basis with each result
A softer scoop can result from several changes in the complete product. Record the formula and process differences before deciding which observation should guide the next trial.
| Check | Record alongside the result | Decision it supports |
|---|---|---|
| Sweetness and finish | Full material identity, actual addition, base, tasting procedure and serving temperature | Does the hardened product meet the taste target? |
| Solids and water | Complete formula by mass; water and solids brought by syrups or other ingredients | Which physical functions still need support? |
| Scoopability or hardness | Defined method, hardening and storage history, sample temperature and time before testing | Is the result suitable for the intended serving condition? |
| Overrun / aeration | Defined same-volume method, net weights and the corresponding mix reference | Is a texture change being confused with a change in air incorporation? |
| Melting behavior | Sample amount and geometry, ambient condition, setup, first drip and subsequent melt observations | Does the melt profile suit the product, without assuming slower is always better? |
| Storage response | Final package, temperature history and agreed checkpoints; texture and ice sensation separately | What changes need another trial or the full validation plan? |
Set the checkpoint schedule for the actual product and intended life. A short frozen sample comparison can guide development; it does not establish commercial shelf life. The existing controlled-trial guide explains the overrun calculation and its limits.
What the latest ice cream study adds
A 2025 International Dairy Journal study compared 15% and 30% sucrose replacement with trehalose, erythritol and isomalt in Fiordilatte ice cream. Overrun, melting, emulsification stability and sensory results changed with the ingredient and level. It supports a control-plus-candidates sample matrix, not a fixed replacement formula. Source: https://doi.org/10.1016/j.idairyj.2025.106339
A published example keeps the replacement mass visible
The FSA Northern Ireland / CAFRE suggested 10%-fat base-mix example changes several ingredients together. It illustrates a system change, not an isolated sweetener comparison.
| Selected recipe ingredient | Control (g/100 g mix) | Reformulated (g/100 g mix) |
|---|---|---|
| Sugar | 9.80 | 4.30 |
| Dextrose | 4.20 | 2.50 |
| Fructose | Not listed (dash) | 3.00 |
| Oligofructose | Not listed (dash) | 4.20 |
Technical references and boundary
The FSA Northern Ireland / CAFRE guidance was published in June 2026 and applies to Northern Ireland; its suggested recipes do not authorise a different product or market. The IFST sugar reformulation brief discusses the technical roles that need attention when changing sugars. Use these references to plan comparisons, then review the actual formula, label and intended market separately.
Compare allulose candidates at a defined serving condition
For a frozen dessert comparison, document the allulose form, full formula, preparation method, freezing conditions and intended serving temperature. Account for water and solids when comparing syrup with powder. Decide whether the first trial holds the formula conditions fixed or adjusts sweetness with another ingredient, and record that choice; otherwise a change in sweetness can become confused with several simultaneous formulation changes. Compare hardness or scoopability at the same defined serving condition, taste using a consistent procedure, and observe melt behaviour with the same sample quantity and setup. Record each result separately against the desired product profile. A slower melt is not automatically a better product, and a preferred taste does not prove storage performance. Use the results to choose the next controlled trial rather than inferring the required allulose level from one observation.
Common questions
Can monk fruit or stevia replace sugar in ice cream?
They can supply sweetness at low use levels, while the complete formula still needs to address body, solids, water and freezing behavior. Compare the actual extract or blend in the hardened product against your reference; matching the sweetness of a liquid mix does not establish the same frozen texture.
Discuss the frozen-dessert sample directionWhy can two equally sweet frozen desserts scoop differently?
They may differ in their full composition, water and solids balance, aeration, process or temperature history. Record those conditions with the hardness or scoopability result rather than assigning the difference to the sweetener name alone.
Discuss the frozen-dessert sample directionDoes the recipe chart prove a sugar-free product or a better blend?
No. It reproduces selected ingredient quantities from a published example, with other recipe context identified. It is not a total-sugars analysis, a performance trial or a finished-pack claim assessment, and it does not rank Hengliyuan blends.
Discuss the frozen-dessert sample directionShould I choose the candidate with the slowest melt?
Use the melt profile that fits the intended product and serving situation. Keep the setup and sample basis consistent and record first drip separately from later melting. Judge those results alongside taste, scoopability and storage texture.
Discuss the frozen-dessert sample directionCan the sample direction be allulose-based or erythritol-free?
State the intended market, ingredient restriction and frozen format. An allulose direction needs the actual ingredient and use route reviewed for that market. Excluding erythritol still leaves sweetness, solids and texture to resolve in the full formula; it does not establish a label claim.
Discuss the frozen-dessert sample directionWhat can I send before the formula and quantity are final?
Begin with the frozen dessert type, final sales market and the problem or product target. Add known base, process, ingredient and package details. We can clarify a sample direction before finalising the specification, quantity and quotation details.
Discuss the frozen-dessert sample directionIs there a fixed allulose-to-monk-fruit ratio for ice cream?
No universal ratio follows from the ingredient names or a practitioner’s rating. Select the allulose addition within the full solids and freezing system, then adjust the identified monk fruit extract toward the sweetness target and record its actual dose. Validate the hardened, stored product.
Discuss the frozen-dessert sample directionDoes V50 mean the amount of monk fruit in the blend?
No. Confirm what the supplied extract grade means on its specification and COA, including the Mogroside V assay and method. That grade designation does not tell you the percentage of extract to add to an allulose blend or to the final ice cream.
Discuss the frozen-dessert sample directionIs erythritol always unsuitable for ice cream?
No. It can create cooling and crystallization concerns, and an erythritol-free brief may be appropriate. Published supplier work also describes useful erythritol-containing systems. Decide within the actual formula, addition, process and storage conditions rather than treating either a low rating or a successful example as universal.
Discuss the frozen-dessert sample directionHow many liters of ice cream can one kilogram of sweetener make?
It depends on the actual sweetener addition and the finished product’s density and overrun. Calculate first on the complete recipe by mass, then use the measured finished yield to express cost per liter. A kilogram of blend and a kilogram of high-intensity extract are different purchasing units.
Discuss the frozen-dessert sample directionCan a Ninja CREAMi example prove a commercial ice-cream formula?
It can suggest a material or sensory direction to test. Record the appliance and processing conditions, then assess the commercial formula under its own freezing, hardening, serving and storage conditions. Freshly processed texture does not establish performance after distribution or refreezing.
Discuss the frozen-dessert sample directionDoes using allulose and monk fruit make ice cream no added sugar?
The sweetener pairing alone does not establish that claim. Review the complete formula, including other sources of sugars, the finished product and the rules of the intended market. No refined sugar, no added sugar and sugar-free describe different questions.
Discuss the frozen-dessert sample directionSee trial directions for this application
