Most comparisons of GLP-1 pills against injections end up comparing different drugs: an oral orforglipron against an injected tirzepatide, say, which are two molecules from two development programmes and cannot separate the route from everything else. There is one exception, and it is the only clean test available. Semaglutide is approved both as a weekly injection and as a daily tablet. Same molecule, same manufacturer, two delivery routes, both documented in FDA labels.
Read side by side, those labels make the pill-versus-injection question much less a matter of preference than it usually sounds.
89% against 1%
The number that governs everything else sits in section 12.3 of each label.
Injected under the skin, the absolute bioavailability of semaglutide is 89%. Nearly all of the dose reaches the bloodstream.
Swallowed, the absolute bioavailability of the same molecule is 0.4% to 1% for Rybelsus at 3, 7 and 14 mg, and 1% to 2% for the Ozempic tablet strengths. Between 99% and 99.6% of the tablet never arrives.
This is not a manufacturing failure. It is what the digestive tract is for. Semaglutide is a peptide, and peptides are what stomach acid and digestive enzymes exist to dismantle, which is why injection was the only route available for this entire drug class for years. The tablet works around the problem rather than solving it: it is co-formulated with an absorption enhancer called SNAC, and — as the label states explicitly — absorption predominantly occurs in the stomach, not further down in the intestine where most oral drugs are taken up. A small window, a small fraction, deliberately engineered.
What that costs in drug
Follow the arithmetic and the scale of the workaround becomes visible.
| Injected | Oral | |
|---|---|---|
| Dose | 2.4 mg | 25 mg |
| Frequency | once weekly | once daily |
| Drug per week | 2.4 mg | 175 mg |
| Bioavailability | 89% | 0.4–1% |
| Trial weight change | −14.9% (week 68) | −13.6% (week 64) |
Roughly 73 times as much semaglutide per week, to land in approximately the same place. Those two trial results come from separate studies four weeks apart in duration — the Wegovy label's Study 1 and OASIS 4 — so the small difference between them should not be read as a ranking. The quantity gap, by contrast, is not an estimate. It is the dose on the carton.
This is also the part with consequences beyond the individual patient. A pill that requires seventy-odd times the active ingredient of an injection needs that ingredient manufactured, and peptide manufacturing capacity is the constraint this entire market has been running into. The convenience of the tablet is purchased with supply.
The fasting window is the mechanism, not advice
Because absorption happens in the stomach through a narrow window, the conditions in that stomach determine the dose. The label's instructions are correspondingly exact: take the tablet on an empty stomach in the morning, with no more than 4 ounces of water and no other liquid, and wait at least 30 minutes before any food, any other drink, or any other oral medication.
The label also shows its working. In a 10-day study in healthy subjects, participants took the tablet with either 50 mL or 120 mL of water and then fasted for 15, 30, 60 or 120 minutes. Water volume made no clinically significant difference. The fasting window did: absorption was higher after a longer post-dose fast.
That is why the daily routine is best understood as part of the delivery system. A coffee twenty minutes after the tablet is not a small deviation from a recommendation; it changes how much drug gets in. The weekly injection carries no equivalent timing rules — the label notes that similar exposure was achieved in the abdomen, the thigh or the upper arm, and nothing about the surrounding hours matters.
What does not change between the routes
Two things stay the same, and both are routinely misunderstood.
The duration is identical. It is natural to assume a daily tablet must be shorter-acting than a weekly shot, and it is not. Peak concentration arrives about an hour after the tablet against one to three days after the injection, but the elimination half-life is approximately one week by either route, and both labels state the molecule remains in circulation for about five weeks after the last dose. Daily dosing here compensates for poor absorption, not for fast clearance.
The pharmacology is the same, because the molecule is the same. The receptor it binds, the gastric emptying it slows, the contraindications, the boxed warning about rodent thyroid C-cell tumours — none of these belong to a needle. A reader hoping that swallowing the drug produces a gentler version of it will not find that in the documents.
Choosing between them
The genuine trade is narrower than the marketing on either side suggests, and it is not mainly about effectiveness.
The injection asks for one act of nerve a week and then makes no further demands on the day. The tablet asks for no needle and a fixed morning: fasted, plain water, half an hour before coffee, breakfast or any other pill — every day, indefinitely, since the absorption depends on it. For somebody who takes a thyroid tablet or a morning blood-pressure medicine, that 30-minute exclusion has to fit alongside instructions those drugs carry too.
Price cuts across this, and the two routes are no longer straightforwardly cheaper or dearer than one another; our GLP-1 pill pricing page sets out what each approved tablet costs at each dose and how that compares with the pens, and the cost hub does the same across the whole category. For the other approved oral option, which is a small molecule rather than a peptide and therefore does not face the absorption problem described here at all, see our orforglipron and tirzepatide comparison.
What no source yet offers is the trial that would settle the route question directly: the same molecule, the same participants, randomised to tablet or pen. Until somebody runs it, the 89% and the 1% are the most honest summary of the difference there is.