High-Intensity Sweeteners Compared: Sucralose, Aspartame, Acesulfame-K and Stevia

The four high-intensity sweeteners in this comparison are not interchangeable. Sucralose (~600x sugar) and acesulfame-K are heat-stable and survive baking and hot fill; aspartame (~200x) is the cleanest-tasting but degrades under heat and acid and carries a mandatory phenylalanine declaration; steviol glycosides (stevia) are the only one of the four that is plant-derived and label-friendly, but they bring a bitter, licorice aftertaste that usually has to be masked or blended away. Choose by potency, thermal and pH stability, taste profile and acceptable daily intake — this guide sets all four side by side.

One framing point first: high-intensity sweeteners are sold by sweetness equivalence, not by weight, so a “200x” sweetener delivers the same perceived sweetness as sugar at roughly 1/200th the dose. That changes how you cost, dose and blend them — and it is why blends, not single sweeteners, dominate real beverage and confectionery formulations. We cover the economics of combining them in the companion guide on sweetener blends and synergy.

The four sweeteners at a glance

Potency from FDA high-intensity sweeteners and Niran Bio comparison. ADI values: FDA aspartame and other sweeteners; JECFA aspartame ADI 40 mg/kg from WHO July 2023; JECFA sucralose 15 mg/kg and steviol glycosides 4 mg/kg as cited by FDA. ADIs differ between authorities — see the ADI section.**

The two ~200x sweeteners (aspartame, acesulfame-K) and the ~600x sucralose are synthetic; stevia is extracted from the Stevia rebaudiana leaf (or its glycosides produced by enzyme conversion or fermentation). That split — synthetic vs plant-derived — drives most clean-label decisions, while heat stability and taste drive most technical ones.

Sucralose (E955)

Sucralose is a chlorinated derivative of sucrose, roughly 600 times sweeter than sugar — the most potent of the four (potency, Niran Bio).

• Heat stability. Good. Sucralose stays sweet at the high temperatures of baking and hot processing, which makes it usable where aspartame is not (heat stability, elchemy). A practical caveat: at very high bake temperatures and long hold times, sucralose can begin to break down, so validate it in the actual process rather than assuming unlimited thermal headroom.
• pH stability. Good across the typical food and beverage pH range, including acidic carbonated drinks — a key reason it is a soft-drink workhorse.
• Taste. Clean and sugar-like with a slight onset delay and lingering tail; generally low bitterness at use levels, though it can show a faint off-note at very high concentrations.
• ADI. FDA sets the sucralose ADI at 5 mg/kg bw/day; JECFA sets it higher at 15 mg/kg bw/day (FDA). For a 60 kg adult the FDA figure is about 300 mg/day. The EU’s most recent re-evaluation maintained its existing position on sucralose with no change to the ADI (EU sucralose re-evaluation, Food Safety Magazine).

Sucralose’s combination of high potency, heat and acid stability makes it the most broadly usable single sweetener of the four — which is also why it anchors so many blends.

Aspartame (E951)

Aspartame is a methyl ester of the dipeptide aspartyl-phenylalanine, about 200 times sweeter than sugar, and widely regarded as the cleanest, most sugar-like taste of the synthetic high-intensity sweeteners (potency, FDA).

• Heat stability. Poor. Aspartame is not heat stable and degrades under prolonged high temperature, which limits or rules it out of baking and many hot-fill processes (heat stability, elchemy).
• pH stability. Sensitive. Aspartame loses sweetness over time in acidic conditions, which constrains shelf life in low-pH beverages — a reason it is often blended with acesulfame-K rather than used alone in carbonated drinks.
• Taste. The benchmark for “clean” — sugar-like with little bitterness — which is why it survives in blends despite its stability limits.
• Phenylalanine declaration. Aspartame is metabolised to phenylalanine, aspartic acid and methanol. Products containing it must carry a phenylalanine declaration so people with phenylketonuria (PKU) can identify it; in the US this warning is mandatory and in the EU a comparable statement applies (PKU labelling, FDA; 21 CFR 201.21, eCFR). This is a labelling fact, not a health claim — the declaration is required regardless of dose.
• ADI and 2023 review. FDA’s ADI is 50 mg/kg bw/day; JECFA’s is 40 mg/kg bw/day. In July 2023 the WHO process produced two parallel outputs: IARC classified aspartame as “possibly carcinogenic to humans” (Group 2B) on limited evidence, while JECFA reviewed the data and kept the ADI unchanged at 40 mg/kg (WHO July 2023; IARC summary). Group 2B is a hazard classification, not a statement of risk at dietary levels; we report the regulatory status factually and make no health claim either way.

Aspartame remains valuable where taste quality matters and the product is neither baked nor strongly acidic — and as a blend component where another sweetener covers its stability gap.

Acesulfame-K (E950)

Acesulfame potassium (Ace-K) is a synthetic sweetener about 200 times sweeter than sugar — as sweet as aspartame, roughly two-thirds as sweet as saccharin, and about one-third as sweet as sucralose (potency, Niran Bio).

• Heat stability. Excellent — stable under heat and under moderately acidic or basic conditions, which makes it suitable for baking and for products needing a long shelf life (heat stability, elchemy).
• pH stability. Good across the beverage range, including hot and acidic systems — which is exactly the weakness it covers for aspartame.
• Taste. Sharp, fast sweetness onset but a slight bitter or metallic tail at higher use levels. This is why Ace-K is rarely used alone; it is almost always blended.
• ADI. FDA’s ADI is 15 mg/kg bw/day (FDA). The EU re-evaluated Ace-K and raised its ADI to 15 mg/kg bw/day, up from the previous 9 mg/kg, based on the highest dose without adverse effects in a chronic study (EFSA re-evaluation, Food Safety Magazine; EFSA re-evaluation opinion, PMC).

Ace-K’s role is structural rather than solo: its heat and acid stability and its synergy with sucralose and aspartame make it a near-universal blend partner.

Steviol glycosides (Stevia, E960)

Stevia is the outlier — the only plant-derived sweetener of the four. The sweet molecules are steviol glycosides extracted from the Stevia rebaudiana leaf, with potency in the ~200–300x range depending on the glycoside (potency context, FDA media).

• Glycoside matters more than “stevia”. Early stevia products relied on rebaudioside A (Reb A) and stevioside, which carry the strongest bitter and licorice notes. Newer grades feature rebaudioside M (Reb M) and Reb D, produced by enzyme-catalysed bioconversion or fermentation, which taste markedly cleaner and sugar-like (Reb M / enzyme and fermentation routes, EFSA via TraceOne; EFSA steviol glycosides opinion 2024). Specifying “stevia” is not enough — specify the glycoside and purity.
• Heat and pH stability. Generally good and suitable for many baked and hot applications, though high-Reb-A grades can show sensory changes under extreme processing; the cleaner Reb M grades are more robust.
• Taste. The defining challenge: a bitter, licorice-like aftertaste with a long tail, strongest in stevioside and Reb A (off-taste of steviol glycosides, Niran Bio). This is masked by switching to Reb M, by masking flavours, or by blending — a blend of Ace-K and sucralose, for example, can mask off-notes while raising overall sweetness (blending to mask off-taste, Niran Bio).
• ADI. JECFA sets the steviol glycosides ADI at 4 mg/kg bw/day, expressed as steviol equivalents — the figure FDA also references (FDA). The EU in 2024 evaluated a proposed modification of the steviol glycosides ADI (from 4 toward 6 or higher, as steviol equivalents) and authorised additional enzyme-/fermentation-produced glycosides (EFSA 2024 opinion). Confirm the current figure for your target market at time of formulation.

Stevia is the choice when a “natural”/plant-derived positioning is required — but the glycoside grade decides whether it tastes acceptable on its own or needs help.

The bulk problem: what high-intensity sweeteners do not replace

A point procurement repeatedly underestimates: high-intensity sweeteners replace sugar’s sweetness but not its bulk. Sugar at 8–12% of a beverage or much higher in a confection contributes body, mouthfeel, freezing-point depression, browning, preservation and structure. Dose a sweetener at 1/200th or 1/600th of that and you remove all of it. The consequences land downstream:

• Beverages lose mouthfeel and can taste “thin”; formulators add bulking and mouthfeel agents, or accept the difference.
• Baked goods lose browning, volume and moisture retention — sugar is a structural ingredient there, not just a sweetener, which is why sugar reduction in bakery is far harder than in drinks.
• Confectionery loses the structural sugar matrix entirely, so a high-intensity sweetener alone cannot make a hard candy or a fondant; a bulk sweetener (a polyol such as maltitol or isomalt) carries the structure while the high-intensity sweetener tops up the sweetness.

This is why high-intensity sweeteners are usually one component of a sugar-reduction system, not a one-for-one swap. Maltodextrin and other carbohydrate bulking agents often fill the body gap; we cover their selection in the additives cluster. The takeaway for sourcing: when a customer asks to “replace the sugar”, clarify whether they need sweetness, bulk, or both — the answer decides whether one sweetener is enough or a system is required.

A worked dosing example

Sweetness equivalence makes the dosing arithmetic simple and explains the cost case. To match the sweetness of 100 g of sugar:

Illustrative, based on the potency figures above; real use levels depend on the matrix, the sweetness target and synergy in a blend. These are sweetness-equivalence estimates, not formulation instructions.

Two practical reads follow. First, the per-gram price of a high-intensity sweetener looks high next to sugar, but the in-use cost is a fraction because so little is dosed — the comparison that matters is cost-per-unit-sweetness, not cost-per-kilo. Second, because doses are so small, weighing accuracy and dispersion become real production issues; sub-gram dosing per batch is unforgiving of poor mixing, which is one reason pre-made blends with carriers are common.

Heat and pH stability — the technical decision

If your product is baked, retorted or hot-filled, stability narrows the field fast:

Stability behaviour from elchemy and Niran Bio formulation guidance.

A documented practical rule: for hot beverages such as tea or coffee, a blend of acesulfame-K and sucralose gives better sweetness stability at neutral pH and high temperature than aspartame used alone (hot-beverage blend, Niran Bio). That is the single most common reason aspartame appears in blends rather than solo in shelf-stable acidic drinks.

Taste, aftertaste and why blends win

No single high-intensity sweetener reproduces the sweetness curve of sugar — each has a different onset, peak and tail, and most carry an off-note:

• Aspartame is the cleanest but unstable.
• Sucralose is clean with a slight delay and lingering tail.
• Ace-K has a fast onset but a bitter/metallic tail.
• Stevia (Reb A) has a bitter, licorice aftertaste; Reb M is much cleaner.

Combining sweeteners produces synergy — each masks the other’s off-note, and the blend can be sweeter than the sum of its parts, which also cuts cost (synergy and mutual masking, Niran Bio). Ace-K plus sucralose is the classic pairing; aspartame is added where taste quality justifies its stability limits; stevia is blended in for a plant-derived story while another sweetener carries the load and masks the licorice. We work through the cost and curve mechanics in sweetener blends and synergy, and the interaction with flavour, pH and carbonation in beverage flavour systems.

Choosing by product category

Translating the properties above into a first-pass choice by application:

• Carbonated soft drinks (low pH, ambient, long shelf life). Acesulfame-K plus sucralose is the durable backbone; aspartame can be added for taste roundness but should not carry the load alone because it fades on the acidic shelf. Stevia (Reb M) is the plant-derived route, usually blended.
• Still and functional drinks. Similar logic, with more room for stevia-forward blends where a “natural” claim is the selling point and the pH is less aggressive.
• Hot beverages and premixes (tea, coffee, instant). Acesulfame-K/sucralose blends win on heat and neutral-pH stability over aspartame alone.
• Bakery. Sucralose and acesulfame-K survive the oven; aspartame does not. Remember the bulk problem — sugar reduction in bakery needs a structural strategy, not just a sweetener swap.
• Dairy and chilled desserts. Cold chain relaxes thermal constraints; choice is driven by taste cleanliness and any clean-label requirement, so stevia (clean grades) and sucralose feature.
• Tabletop and sachets. Single-serve dosing accuracy dominates; carriers and blends manage the micro-dose, and aspartame’s clean taste is attractive where the product is not heated.
• Confectionery. A bulk sweetener (polyol) plus a high-intensity top-up; the high-intensity choice follows the heat and acid profile of the specific sweet.

This is a starting grid, not a recipe — the final choice is validated in the actual matrix, because flavour system, pH, temperature and carbonation all shift the sweetness curve, as covered in beverage flavour systems.

ADI: read the right authority

ADI values differ by regulator, and using the wrong one can misstate your headroom:

All in mg/kg bw/day. FDA values from FDA; JECFA aspartame from WHO July 2023; EFSA Ace-K from Food Safety Magazine; EFSA steviol from EFSA 2024.

For the Egyptian market, where the absence of a specific national standard typically means Codex Alimentarius / JECFA figures apply, the JECFA ADI is usually the right reference — but always confirm the maximum permitted level for your specific food category against the current Egyptian requirement rather than working from a foreign limit. We make no health claim about any of these sweeteners; ADIs are stated as regulatory facts.

Identity and labelling: get the names right

Each sweetener carries multiple identifiers, and getting them right on the spec and the label avoids customs and compliance friction:

• Sucralose — E955, INS 955.
• Aspartame — E951, INS 951; triggers the mandatory phenylalanine declaration.
• Acesulfame-K (acesulfame potassium) — E950, INS 950.
• Steviol glycosides — E960, with sub-categories (E960a–d) distinguishing leaf-extract, enzyme-modified and fermentation-derived glycosides under the EU framework (E960 sub-categories, EFSA via TraceOne).

On an Egyptian label the additive must be declared by its function and name or number, and aspartame’s phenylalanine statement must appear. For the underlying reading of E-numbers versus INS versus CAS identity across regulators, see the additives identity guide in the same cluster. Specify the identifier and the grade on the purchase order; “stevia” or “sweetener” alone is not a buyable specification.

Storage, handling and dispersion

The sweeteners differ in handling as much as in taste:

• Hygroscopicity and caking. Several high-intensity sweeteners pick up moisture and cake; store sealed, cool and dry — relevant for Egyptian summer warehousing — and confirm anti-caking carriers where supplied.
• Dispersion at micro-dose. Because so little is dosed, uniform dispersion is a genuine production risk. Pre-dilution, a carrier, or a ready-made blend solves the weighing-accuracy and mixing problem that sub-gram-per-batch dosing creates.
• Stability in storage. Aspartame is the most fragile in storage as well as in process, losing sweetness over time especially in moisture or acid; the others are more robust. Rotate stock and respect shelf life rather than assuming indefinite potency.

How Innovote sources this

Sweeteners are easy to buy badly because the spec hides the decisions that matter — glycoside grade, blend ratio, particle size and the certificate scope. We pin them down.

• We start from the application stress, not the sweetener name: baked, hot-filled or acidic products eliminate aspartame-alone before price is even discussed.
• We specify the grade, not just the molecule. For stevia that means the glycoside (Reb A vs Reb M) and purity; for the synthetics it means particle size and any anti-caking carrier that affects dispersion.
• We default to blends where they win — Ace-K/sucralose for stability and cost, with aspartame or stevia added for taste or positioning — and validate the sweetness curve in your actual matrix.
• We collect the CoA and identity documents — E-number/INS identity, purity, heavy metals, microbiology — and we confirm the phenylalanine declaration is in place wherever aspartame is used.
• We map the regulatory basis to Egypt — typically Codex/JECFA in the absence of a national standard — and flag the maximum permitted level for the food category so the formulation clears at the border.

Innovote is an Egyptian sourcing partner; our sweetener offers are stated as compliant with / meets the requirements of the relevant specification, with certificates and specifications available on request. We do not describe any sweetener as “approved” without a documented basis, and we make no health or medical claims — the IARC and ADI figures above are reported as regulatory status, not as advice.

FAQ

Which high-intensity sweetener is the sweetest?
Sucralose, at roughly 600x sugar. Aspartame and acesulfame-K are each around 200x, and steviol glycosides are roughly 200–300x depending on the glycoside (FDA; Niran Bio).

Which sweeteners can I use in baking?
Sucralose, acesulfame-K and most stevia grades tolerate baking heat; aspartame does not and is generally avoided in baked and hot-fill products because it degrades under prolonged heat (elchemy).

Why does stevia taste bitter, and how is it fixed?
The bitter, licorice aftertaste comes from steviol glycosides such as stevioside and Reb A. It is reduced by switching to cleaner-tasting Reb M grades (made by enzyme conversion or fermentation), by masking flavours, or by blending with other sweeteners that cover the off-note (Niran Bio; EFSA 2024).

Did the 2023 aspartame review change its legal status?
No. In July 2023, IARC classified aspartame as “possibly carcinogenic to humans” (Group 2B, a hazard category on limited evidence), while JECFA reviewed the same data and kept the acceptable daily intake unchanged at 40 mg/kg bw/day. Aspartame remained authorised (WHO July 2023).

Why are sweeteners usually sold and used as blends?
Because no single one matches sugar’s full sweetness curve, and combining them creates synergy — each masks the other’s off-note and the blend is often sweeter than the sum of its parts, which lowers cost. Ace-K plus sucralose is the classic pairing (Niran Bio). See sweetener blends and synergy.

Which ADI applies to imports into Egypt?
In the absence of a specific Egyptian standard, Codex Alimentarius / JECFA figures generally apply, so the JECFA ADI is usually the right reference rather than the FDA value. Always confirm the maximum permitted level for your specific food category against the current Egyptian requirement.

Source it with Innovote

Tell us the product, the process (baked, hot-fill or cold/acidic) and whether you need a plant-derived positioning, and we will come back with the right sweetener or blend, the glycoside grade where stevia is involved, MOQ, lead time and a landed-cost path into Egypt — with certificates and specs on request.

Related reading: Food Additives & Functional Ingredients hub · Sweetener blends and synergy · Beverage flavour systems: matching flavour to pH, sweetener and carbonation

Byline: Innovote Trade Desk. Compliance note: capability statements are phrased as “compliant with / meets the requirements of / certificates and specifications available on request.” ADI and IARC figures are reported as regulatory status, not health advice; no health or medical claims are made. Confirm all regulatory figures and maximum permitted levels against the current Egyptian/Codex requirements at time of import.