Education Center
Semaglutide vs Tirzepatide: Built from Four Labels and Two Head-to-Head Trials
Most peptide comparisons are assembled by inference — one compound’s trials read against another’s, in different populations, at different endpoints. This pair is the exception: both carry FDA-approved labels, and two randomised trials have measured them against each other directly and published the result. This page is built from those documents and says plainly where they stop applying. Research-use education only: no dosing recommendations, no injection instructions, no treatment protocols, no medical advice.
The short answer
Tirzepatide beat semaglutide on the primary endpoint of both randomised head-to-head trials that have reported — and in both trials semaglutide was capped below the maximum its labeling now describes. Those two sentences have to travel together. The first is what nearly every summary reports; the second is what almost none of them do, and it is the difference between a result and a headline. The framing that usually opens this comparison — one receptor against two — is accurate and stated on both labels, but receptor count is the least interesting thing here: it is the one claim requiring no evidence beyond a definition. What makes this pair worth a page is that the harder questions have documents behind them.
Four labels, not two
Neither compound has one label. Each is approved under two brand names, for different populations, with different ceilings, so comparing “semaglutide” to “tirzepatide” without saying which label is meant is wrong about half the time. Every row comes from the manufacturer’s Structured Product Label on DailyMed. No starting dosage or escalation schedule is reproduced — only the labeled ceiling, because the trial section below turns on it. That ceiling is label text about a prescribing clinician’s decision, not guidance to anyone reading this.
| Label | Approved indications, as stated | Labeled maximum, as stated |
|---|---|---|
| Ozempicsemaglutide injection · Novo Nordisk | Adults with type 2 diabetes: glycaemic control; major adverse cardiovascular event risk reduction with established cardiovascular disease; and, with chronic kidney disease, risk reduction for sustained eGFR decline, end-stage kidney disease and cardiovascular death | 2 mg once weekly, stated as the maximum recommended dosage |
| Wegovysemaglutide injection and tablets · Novo Nordisk | Injection: cardiovascular event risk reduction with established cardiovascular disease and obesity or overweight; weight reduction and long-term maintenance in adults, in patients aged 12 and older with obesity, and in adults with overweight plus a weight-related condition; noncirrhotic MASH with F2 to F3 fibrosis, under accelerated approval. Tablets: the cardiovascular and weight indications, adults only | Injection: 2.4 mg once weekly maintenance, 7.2 mg stated as the maximum for weight reduction in adults. Tablets: 25 mg once daily |
| Mounjarotirzepatide injection · Eli Lilly | Glycaemic control in adults and patients 10 years and older with type 2 diabetes; major adverse cardiovascular event risk reduction in adults with type 2 diabetes at high risk for those events | 15 mg once weekly in adults; 10 mg in patients 10 years and older |
| Zepboundtirzepatide injection · Eli Lilly | Weight reduction and long-term maintenance in adults with obesity, or overweight plus at least one weight-related condition; moderate to severe obstructive sleep apnoea in adults with obesity | 15 mg once weekly, stated as the maximum for both indications |
They are not competing for the same ground everywhere. Only semaglutide carries a kidney-outcome indication and a MASH indication, the latter under accelerated approval and so conditional on a confirmatory trial. Only tirzepatide carries an obstructive sleep apnoea indication. Both reach children, but not the same children: Wegovy injection covers obesity from age 12, Mounjaro covers type 2 diabetes from age 10 — a distinction most side-by-side summaries collapse into “adults only.”
The two head-to-head trials
SURPASS-2 (NCT03987919) is the type 2 diabetes comparison: a 40-week phase 3 trial sponsored by Eli Lilly, randomising 1,879 adults inadequately controlled on metformin 1:1:1:1 to tirzepatide 5, 10 or 15 mg or to semaglutide 1 mg weekly. Mean baseline HbA1c was 8.28%, mean weight 93.7 kg. Its masking takes two words rather than one: the registry records no masking, while the Mounjaro label describes an open-label trial that was double-blind with respect to tirzepatide dose assignment. Nobody was blinded to which compound they received; the three tirzepatide arms were blinded against each other. A registry record proves only that somebody filed a form, so it is worth saying how this one was corroborated — the same trial appears in section 14.3 of the FDA-approved Mounjaro labeling, same design, same patient count, same comparator. Reaching approved labeling means the underlying data passed regulatory review: a stronger check than a registry entry, and the one worth applying to any cited trial.
| Arm | HbA1c change, week 40 | vs comparator | Body weight change | vs comparator |
|---|---|---|---|---|
| Tirzepatide 5 mg | -2.09 points | -0.23 points | -7.8 kg | -1.6 kg |
| Tirzepatide 10 mg | -2.37 points | -0.51 points | -10.3 kg | -4.1 kg |
| Tirzepatide 15 mg | -2.46 points | -0.60 points | -12.4 kg | -6.2 kg |
| Semaglutide 1 mg | -1.86 points | — | -6.2 kg | — |
Those are the results posted on the registry record, with the difference columns computed from them rather than transcribed. The paper reaches the same conclusion — tirzepatide noninferior and superior on mean HbA1c change to 40 weeks — from a different set of figures, which the next section takes up. Adverse events were predominantly gastrointestinal and mostly mild to moderate in both groups: nausea in 17 to 22% of tirzepatide patients against 18% of semaglutide patients, diarrhoea 13 to 16% against 12%, vomiting 6 to 10% against 8%, serious adverse events 5 to 7% against 3%.
SURMOUNT-5 (NCT05822830) is the obesity comparison, and the cleaner of the two: a 72-week open-label phase 3b trial, again sponsored by Eli Lilly, randomising 751 adults with obesity but without type 2 diabetes 1:1 to the maximum tolerated dose of tirzepatide (10 or 15 mg) or of semaglutide (1.7 or 2.4 mg). Least-squares mean percent weight change at week 72 was −20.2% with tirzepatide (95% CI −21.4 to −19.1) against −13.7% with semaglutide (95% CI −14.9 to −12.6), a gap of 6.5points; waist circumference fell −18.4 cm against −13.0 cm, a gap of 5.4cm. Both reached P<0.001.
One trial, three documents, two different analyses
SURPASS-2 exists in three primary documents — the New England Journal paper, section 14.3 of the Mounjaro label, and the posted results on its registry record. All three describe the same 1,879 patients over the same 40 weeks. They do not print the same numbers, and the split is not three ways.
| Document | HbA1c: semaglutide 1 mg, tirzepatide 15 mg | Body weight: same two arms | Analysis population, as the document describes it |
|---|---|---|---|
| NEJM 2021, Frías et al. | -1.86 and -2.30 points | Treatment differences only: -1.9, -3.6 and -5.5 kg | The abstract does not describe its analysis population |
| Mounjaro labeling, section 14.3 | -1.9 and -2.3 | -5.7 kg against -11.2 kg | Modified intent-to-treat; missing week-40 data imputed by multiple imputation with retrieved dropout |
| Posted results, NCT03987919 | -1.86 and -2.46 points | -6.2 kg against -12.4 kg | Data after rescue medication or premature discontinuation of study drug excluded |
The paper and the label agree: their HbA1c figures differ only by rounding, and the paper’s body-weight treatment differences — −1.9, −3.6 and −5.5 kg — are exactly the label’s. The registry is the outlier on both endpoints, at −2.46 rather than −2.30 points for tirzepatide 15 mg and −6.2 kg rather than −5.7 kg for semaglutide. None of this is error. Two documents report one pre-specified analysis and the third reports another, and the fourth column is where each says so: the registry drops data recorded after a patient started rescue medication or stopped study drug, while the label carries those patients back in by imputing their missing week-40 values from retrieved dropouts.
One limit belongs here rather than in a footnote: neither the registry record nor the paper’s abstract names its analysis in the vocabulary trial statisticians use for it, so the reading above comes from the population descriptions those documents print, not from a term either applies. That is enough to explain the discrepancy. It is not the same as a document that labels it. What follows is a rule worth carrying into any peptide literature: a number from a trial is not complete until it names the document it came from. Two people can quote SURPASS-2 accurately and disagree by half a kilogram, and neither is wrong.
What the head-to-heads did not test
Four limits bound what these two trials establish. Each is checkable rather than asserted, so the records checked are named.
- Neither trial ran semaglutide at the ceiling its current labeling describes. SURPASS-2 used 1 mg; Ozempic now states 2 mg once weekly as the maximum recommended dosage, and Drugs@FDA records the efficacy supplement to NDA 209637 as approved on 28 March 2022, more than a year after SURPASS-2 completed in February 2021. (That record confirms the date and the supplement type, not the dosage.) SURMOUNT-5 capped semaglutide at 2.4 mg; Wegovy gives 2.4 mg as the recommended maintenance dosage for weight reduction in adults and states 7.2 mg as the maximum. Tirzepatide ran at or beside 15 mg in both. That does not invalidate either result. It bounds it.
- No randomised trial has compared the two on cardiovascular outcomes. The records checked were SELECT (NCT03574597), 17,604 participants, semaglutide 2.4 mg against placebo; SURPASS-CVOT (NCT04255433), 13,299 participants, tirzepatide against an active comparator that is dulaglutide 1.5 mg, not semaglutide; and four large records naming both compounds — NCT06914141, NCT07096063, NCT07619508 and NCT06779929, all registered to Brigham and Women’s Hospital and all observational database analyses. A database emulation is evidence, but not the same evidence as a randomised head-to-head.
- Two is the count of published head-to-heads, not of every trial that randomises to both. Other interventional records do, the largest being NCT06803888 — a phase 4 trial setting bariatric surgery against semaglutide and against tirzepatide in an estimated 140 participants, primary endpoint percent total body weight loss at week 52, listed as active and not recruiting and so reporting nothing. The rest are small physiology, body-composition and exercise studies. SURPASS-2 and SURMOUNT-5 remain the only published head-to-head efficacy trials.
- Both trials were open-label and sponsored by the manufacturer of the compound that won. Stated in both registry records and both papers, and not a scandal — head-to-head trials are almost always funded by someone with a stake in the answer. It belongs beside the result rather than beneath it.
Side by side, from the labels
Every value below is stated in one of the two manufacturer labels. Nothing is inferred, averaged or taken from a secondary source.
| Property | Semaglutide | Tirzepatide |
|---|---|---|
| Receptor description on the label | GLP-1 receptor agonist | GIP and GLP-1 receptor agonist |
| Elimination half-life | Approximately 1 week | Approximately 5 days |
| Time to steady state | 4 to 5 weeks of weekly dosing | 4 weeks of weekly dosing |
| Absolute bioavailability | 89% | 80% |
| Plasma protein binding | Albumin-bound, over 99% | Albumin-bound, 99% |
| Apparent volume of distribution | Approximately 12.5 L | Approximately 10.3 L |
| Fatty-acid modification, as described | C18 fatty diacid at position-26 lysine, via a hydrophilic spacer | C20 diacid at the position-20 lysine, via a linker |
| Primary route of elimination | Metabolism: proteolytic cleavage, then beta-oxidation of the side chain | Proteolytic cleavage, beta-oxidation of the C20 diacid, amide hydrolysis |
| Boxed warning | WARNING: RISK OF THYROID C-CELL TUMORS | WARNING: RISK OF THYROID C-CELL TUMORS |
| Molecular formula | C187H291N45O59 | C225H348N48O68 |
The half-life row is the one that surprises people. The dual-agonist molecule is larger by 699.95Da and is cleared faster. Both carry a fatty diacid and both are almost entirely albumin-bound, but only one label draws the causal line: Ozempic states outright that albumin binding is semaglutide’s main protraction mechanism, while Mounjaro reports the structure and the 99% binding without attributing the half-life to either. Neither label puts duration of action down to slowed renal clearance, and the elimination row is why — both are cleared by metabolism, and intact tirzepatide is not observed in urine or faeces at all. Size and duration of action are not the same axis, so “bigger and more receptors” is not shorthand for “stronger and longer.”
It also shows why the label has to be named: both semaglutide labels state the same one-week half-life, then different consequences — Ozempic, present in the circulation about five weeks after the last dose; Wegovy, about five to seven weeks after a 2.4 mg or 7.2 mg injection.
Molecular identity, verified
Identity is the one property checkable before anything else. Semaglutide is PubChem CID 56843331, formula C187H291N45O59; the Ozempic labeling states a molecular weight of 4113.58 g/mol. Tirzepatide is PubChem CID 156588324, formula C225H348N48O68; the Mounjaro labeling states 4813.53 Da. The 699.95 Da gap is computed from those two figures rather than quoted.
One caveat belongs with those identifiers rather than in a footnote. A PubChem name search for semaglutide resolves to a single record; the same search for tirzepatide resolves to several, all carrying the identical formula. The CID cited above is the one PubChem titles “Tirzepatide” and whose formula and mass match the Mounjaro labeling. A database identifier is a pointer to a description, never evidence about what is in a particular vial.
The labels explain what the extra mass is doing. Ozempic describes semaglutide as a GLP-1 analogue with 94% sequence homology to human GLP-1, modified at position 8 to stabilise it against dipeptidyl-peptidase 4. Mounjaro places tirzepatide on the GIP sequence, with aminoisobutyric acid at positions 2 and 13 and a C-terminal amide; it states no residue count, so none appears here. One is a modified GLP-1, the other a modified GIP engineered to reach the GLP-1 receptor as well — a different starting point, not a variation on the same one.
Where this comparison stops applying
Everything above describes four approved drug products, and the readership for a page like this is largely working with research-grade material. The transfer is partial.
What transfers: identity, conditionally. If a material genuinely matches C187H291N45O59 or C225H348N48O68, the receptor pharmacology on the labels is a property of that molecule, not of the brand on the box. Identity is also the one attribute an independent analytical report can actually establish, which is why what such a document can and cannot prove is worth understanding before relying on one.
What does not transfer: essentially everything else. The half-lives, bioavailabilities, adverse-event rates and both sets of trial results were generated with specific approved presentations, under the fill, excipient, container-closure and stability controls those approvals rest on. A trial result is a property of the whole product and the whole protocol, not of the active molecule in isolation. No registered study compares compounded or research-grade material against an approved product. The records checked were NCT02070510, NCT04143945 and NCT04007107 (pharmacokinetic and injection-site studies), NCT07373834 and NCT07554417 (muscle and dietary outcomes), NCT07588984 and NCT07092618 (behavioural programmes and post-cessation maintenance), and the closest of them, NCT07092605 — sublingual and subcutaneous “microdosed” GLP-1 receptor agonists, but against a sublingual placebo rather than an approved product, and enrolling by invitation with nothing reported.
The reasonable inference is narrow, and stating it narrowly is the point. These trials establish something real about two molecules under two protocols. They establish nothing about any particular vial, and no page — this one included — closes that gap.
Everything on this page is research-use education drawn from labels, registry records and journal reports. Nothing here is a dosing, route or frequency instruction, and nothing here describes what anyone should do.
What Medibact supplies, and what it does not
Medibact supplies USP-grade bacteriostatic water produced in an FDA-registered U.S. facility, for research use only, and publishes research guides including entries for semaglutide and tirzepatide. It does not supply either compound. The reconstitution sections of those guides convert between amount, volume and concentration; they are arithmetic and produce no recommendation about how much of anything to use. Medibact is not affiliated with Novo Nordisk, Eli Lilly or any other manufacturer named here. Ozempic, Wegovy, Mounjaro and Zepbound are named because their FDA labels are the primary documentary record for these two compounds.
Where this sits in the rest of the reference
FAQ
What is the difference between semaglutide and tirzepatide?
Most comparisons lead with receptor count: semaglutide's labeling describes a GLP-1 receptor agonist, tirzepatide's a GIP and GLP-1 receptor agonist. The difference that matters more is documentary — both have FDA-approved labels, and two randomised trials have compared them directly and published the result. Almost every other peptide comparison has to be assembled by inference across separate studies. This one does not.
Has anyone compared semaglutide and tirzepatide head to head?
Twice with published results, both sponsored by Eli Lilly. SURPASS-2 (NCT03987919) randomised 1,879 adults with type 2 diabetes to tirzepatide 5, 10 or 15 mg or semaglutide 1 mg for 40 weeks (Frías et al., N Engl J Med 2021;385:503-515). SURMOUNT-5 (NCT05822830) randomised 751 adults with obesity but without diabetes to the maximum tolerated dose of each for 72 weeks (Aronne et al., N Engl J Med 2025;393:26-36). Other registered trials randomise to both — the largest is NCT06803888 — but none has reported.
Which compound produced the larger reduction in those trials?
Tirzepatide, on the primary endpoint of both. SURPASS-2's paper reports an HbA1c change of -2.30 percentage points for tirzepatide 15 mg against -1.86 for semaglutide 1 mg (difference -0.45, 95% CI -0.57 to -0.32). SURMOUNT-5 reports a least-squares mean weight change at week 72 of -20.2% against -13.7%. Both trials were open-label and funded by the manufacturer of the winning arm.
Did the head-to-head trials test the highest doses on each label?
No, and this is the most important caveat here. SURPASS-2 used semaglutide 1 mg; Ozempic now states 2 mg once weekly as the maximum recommended dosage, approved after that trial ended. SURMOUNT-5 capped semaglutide at 2.4 mg; Wegovy gives 2.4 mg as the recommended maintenance dosage for weight reduction in adults and states 7.2 mg as the maximum. Tirzepatide ran at or beside its labeled ceiling of 15 mg in both. That does not invalidate either result — it bounds it.
Do the two compounds carry the same warnings?
The structures are close to identical. Both the Ozempic and Mounjaro labels open with a boxed warning headed WARNING: RISK OF THYROID C-CELL TUMORS, and both are contraindicated on a personal or family history of medullary thyroid carcinoma, on Multiple Endocrine Neoplasia syndrome type 2, and on serious hypersensitivity. Both warn on acute pancreatitis, hypoglycaemia with insulin secretagogues or insulin, acute kidney injury from volume depletion, severe gastrointestinal reactions, gallbladder disease and pulmonary aspiration under anaesthesia. The wording difference worth noticing is the retinopathy warning: Ozempic's covers diabetic retinopathy complications generally, Mounjaro's is scoped to patients with a history of it.
Which one has the longer half-life?
Semaglutide. Ozempic states an elimination half-life of approximately one week, with the compound present in the circulation for about five weeks after the last dose; Mounjaro states approximately five days. The larger, two-receptor molecule is the shorter-lived one. Both carry a fatty diacid — C18 for semaglutide, C20 for tirzepatide — and both are almost entirely albumin-bound, but only Ozempic states that albumin binding is the protraction mechanism. Mounjaro reports the structure and the binding without attributing the half-life to either.
Does any of this transfer to research-grade material?
Only partly, and not the parts people most want. Identity transfers if the material genuinely matches the published formula. Everything downstream of identity — pharmacokinetics, adverse-event rates, trial results — was generated using specific approved presentations under the controls those approvals rest on. No registered study compares compounded or research-grade material against an approved product; the records checked are named on this page, and the closest (NCT07092605) tests a sublingual preparation against a sublingual placebo and has reported nothing.
What does Medibact supply, and is this medical advice?
Medibact supplies USP-grade bacteriostatic water produced in an FDA-registered U.S. facility, for research use only, alongside its written research guides. It does not supply semaglutide or tirzepatide and is not affiliated with Novo Nordisk, Eli Lilly or any manufacturer named here. This page is label and literature education — not medical advice, not a dosing, route or frequency instruction, and not personal-use guidance.
Sources
- DailyMed SPL set ids — Ozempic adec4fd2-6858-4c99-91d4-531f5f2a2d79; Wegovy ee06186f-2aa3-4990-a760-757579d8f77b; Mounjaro d2d7da5d-ad07-4228-955f-cf7e355c8cc0; Zepbound 487cd7e7-434c-4925-99fa-aa80b1cc776b.
- Frías JP, et al. N Engl J Med 2021;385(6):503-515. PMID 34170647, DOI 10.1056/NEJMoa2107519 — SURPASS-2.
- Aronne LJ, et al. N Engl J Med 2025;393(1):26-36. PMID 40353578, DOI 10.1056/NEJMoa2416394 — SURMOUNT-5.
- Records checked — cardiovascular-outcome absence: NCT03574597, NCT04255433, NCT06914141, NCT07096063, NCT07619508, NCT06779929. Head-to-head count: NCT06803888. Compounded-material absence: NCT02070510, NCT04143945, NCT04007107, NCT07373834, NCT07554417, NCT07588984, NCT07092618, NCT07092605.
- PubChem CID 56843331 and CID 156588324; Drugs@FDA NDA 209637, efficacy supplement 9, approved 28 March 2022.
- Where these documents do not settle a question, the page says so rather than filling the gap.
Educational use only — not medical advice. This guide summarizes information reported in published research and community practice for educational purposes. It is not medical advice and not a recommendation to use any compound. Any doses, schedules, or combinations shown are examples of what has been reported, not instructions for you. Many peptides described here are research compounds that are not FDA-approved for the uses discussed and may be investigational or restricted. Effects, risks, and legal status vary; individual needs and results vary. Consult a qualified, licensed healthcare professional before making any decision. Do not use this content to diagnose, treat, or dose yourself.