Compounded semaglutide is now sold in forms meant to dissolve in the mouth: products held under the tongue, and tablets that disintegrate before they are swallowed. The approved semaglutide tablets work in a very particular way, and that way does not include the mouth. So the reasonable question is whether semaglutide can get into the blood through the lining under the tongue at all, and what has actually been measured when someone tried.
The short answer is that it has been measured in rats and minipigs, and not in people. This page walks through what those studies found and why the approved tablet is not evidence either way. For the approved daily tablet taken for weight, see what the Wegovy pill trials show.
How the approved tablet gets absorbed
Semaglutide is a peptide, and peptides taken by mouth are mostly broken down or simply fail to cross the gut wall. The approved tablets solve this with a co-formulated absorption enhancer, salcaprozate sodium (SNAC), and their label states that absorption happens predominantly in the stomach [1].
Even with that help the numbers are small. The label puts estimated absolute bioavailability at roughly 0.4% to 1% for the 3, 7 and 14 mg tablets, and 1% to 2% for the newer 1.5, 4 and 9 mg tablets, with peak levels about an hour after a dose [1].
The mechanism behind that was mapped in human and dog studies by the drug’s maker. Absorption took place in the stomach, was confined to a small area right next to the surface of the dissolving tablet, and did not happen without SNAC in the same tablet [2]. The enhancer protected the peptide from digestive enzymes by buffering its immediate surroundings, and it raised uptake only briefly.
That is why the label’s instructions read the way they do. It says to swallow the tablet whole and not to split, crush, chew or dissolve it in any solution, to take it with no more than 4 ounces of water, and to wait at least 30 minutes before eating or drinking [1]. The label also reports that absorption was higher with 50 mL of water than with 240 mL, and higher with longer waits after the dose.
Why the mouth is a different problem
The lining under the tongue has real appeal as a route. It is well supplied with blood vessels, and a drug absorbed there skips the stomach and the first pass through the liver [3].
For peptides the same review lists three obstacles: saliva washes the drug away, enzymes in the mouth degrade it, and the tightly packed, lipid-rich lining resists large molecules [3]. Getting consistent peptide absorption from a product held in the mouth remains, in the authors’ word, challenging. The strategies the review covers lean on permeation enhancers, mucoadhesive polymers and multilayer films, each engineered to hold the peptide against the lining for longer.
There is also a question the published work does not answer directly. Whatever dissolves in the mouth and is then swallowed reaches the stomach as saliva, not as an intact tablet sitting against the stomach wall with its enhancer. The stomach absorption described above depended on exactly that arrangement [2], so a mouth-dissolved product should not be assumed to borrow it. That is an inference from the mechanism, not a measured result.
What the rat study measured
The one published pharmacokinetic study of sublingual semaglutide was carried out in Sprague-Dawley rats by researchers employed by a supplier to compounding pharmacies [4]. Rats received semaglutide under the tongue in a proprietary water-free suspension, as a swallowed tablet, or as an injection under the skin.
The doses were not matched. The injection used 0.011 mg/kg, while the sublingual suspension used 1 mg/kg, about 90 times as much, and the swallowed tablet used 1 mg/kg or 20 mg/kg [4]. Semaglutide was detectable in plasma within 2 minutes in every group except the swallowed 1 mg/kg tablet.
At the same 1 mg/kg dose, the area under the curve was 82.53 ng·h/mL under the tongue against 15.08 ng·h/mL by mouth (p=0.004), so in rats the sublingual route beat swallowing [4]. The authors reported relative bioavailability of 0.34% and 0.29% for the sublingual suspensions, made from commercial tablets and from peptide powder, against 0.06% and 0.16% for the swallowed tablets.
Two limits sit alongside that result. It is a single dose in rats, which says nothing about weight or blood sugar in people. And the rat figure cannot be set against the human label’s 0.4% to 1%, since the species, the reference and the method all differ. The authors themselves call it a proof of concept and ask for studies in humans [4].
What the minipig study measured
A second study tested a dissolving film placed under the tongue, loaded with semaglutide and a permeation enhancer, either SNAC or sodium dodecyl sulfate [5]. Forty female Göttingen minipigs, 6 to 7 months old and neither diabetic nor obese, were split between the film, the approved swallowed tablet, the approved weekly injection and placebo.
The film and the tablet were given three times a week and the injection once a week, for 8 weeks with a week of follow-up [5]. Against placebo, the film attenuated weight gain by 24.9% over 9 weeks, with a statistically significant difference in growth trajectory (p=0.013). The authors report that the effect surpassed the swallowed tablet group’s.
These were young animals that were still gaining weight, so the result is slower gain rather than weight loss, and the tablet was not dosed on its labeled daily schedule. The authors name their small group and short duration as limitations [5]. The film is also an engineered delivery system with its own enhancer, which is not the same product as a compounded sublingual preparation.
What has been measured in people
Nothing that has been published. ClinicalTrials.gov lists an early-phase, industry-sponsored study with arms for sublingual semaglutide, injected semaglutide and a sublingual placebo, with an estimated 150 participants [6]. Its record gives an estimated primary completion date of November 2025, was last updated in July 2025, and had no results posted in September 2026.
Its primary outcomes are self-reported pain, mood, sleep, physical activity, heart rate variability and routine blood counts and metabolites [6]. None of them is a blood level of semaglutide, so even a finished version of that study would not show how much of the drug is absorbed under the tongue.
A 2026 review of compounded semaglutide and tirzepatide products reached the same place from the market side. Of 17 single-ingredient products it found, 82% were sublingual and 18% were orally disintegrating tablets, and the authors concluded that whether these forms offer any advantage over the approved oral tablet has not been determined [7]. Only 18% of the products it reviewed reported a beyond-use date and storage conditions.
What this means if you are offered it
The fair summary is that absorption through the mouth has been shown to happen in two animal species, using specific research formulations, at small fractions of the dose given [4] [5]. How much reaches the blood in a person, from any compounded product, has not been measured.
If you are weighing one, two questions are worth putting to the prescriber: which study shows how much semaglutide this particular product delivers to the blood, and how its strength compares with the approved tablet or injection. A precise answer to either would be more than the published record currently holds.
The site tracks sellers of these forms on the GLP-1 tablets and lozenges board, which ranks each on that product rather than on its injection. The same evidence gap exists for another peptide sold under the tongue, covered in whether sermorelin works without an injection. For what compounded status means in law, see are peptides legal.