Where the category stands in May 2026 — what's available, who to trust, what costs what. 22-min read, medically reviewed.
Read the guideAn independent overview of the compounded-GLP-1 telehealth market and the providers we cover — market size, what the FDA changed, and an honest read on each.
View the provider overviewOverview & reviewsPrimers for newcomers and deep dives for the curious. The foundation articles are published; more are in the pipeline.
View all Learn articlesWhat compounded and branded semaglutide and tirzepatide actually cost, from each provider's own published pricing.
Which GLP-1 wins on weight loss, side effects, and cost — the head-to-head SURMOUNT-5 data and what it actually means for choosing.
For obesity without diabetes, the head-to-head SURMOUNT-5 trial settled the long-running question: at maximum tolerated doses over 72 weeks, tirzepatide produced a mean 20.2% body-weight loss versus 13.7% for semaglutide — a 6.5-percentage-point advantage [1]. Both are once-weekly injections with comparable, mostly gastrointestinal side effects [1]. Semaglutide retains the stronger cardiovascular-outcomes evidence [9]. Liraglutide, the older daily agent, now sits well below both.
What follows is an evidence-ranked comparison of the three injectable agents — liraglutide, semaglutide, tirzepatide — that moved chronic weight management from a field where 5% total body-weight loss was a meaningful signal to one where the question is whether 20% is enough. The framing is the one a clinician uses at the keyboard: what does the best available trial show, who is the patient in front of me, and what would change my mind.
Until May 2025 the strongest comparative data was SURPASS-2, a type 2 diabetes trial that pitted tirzepatide against semaglutide 1 mg — a dose nobody uses for obesity [5]. Indirect comparisons from the STEP and SURMOUNT programs suggested tirzepatide produced roughly five to seven additional percentage points of body-weight loss versus semaglutide 2.4 mg [4], but the studies enrolled different populations, ran for different durations, and used different lifestyle interventions. Sophisticated readers discounted the gap. Less sophisticated readers, including many prescribers, took the numbers at face value.
SURMOUNT-5 closed that gap by enrolling people with obesity and no diabetes, randomizing them to maximum tolerated doses of either drug, and running both arms for 72 weeks under the same protocol [1]. The result is the first apples-to-apples weight-loss comparison of the two agents at the doses people actually take. It's also the result that should drive prescribing now, where both agents are accessible and price-equivalent.
Semaglutide and liraglutide bind a single incretin receptor, GLP-1R, expressed in pancreatic beta cells, the gut, the vagal afferents, and — most relevant to weight loss — the hypothalamic and brainstem nuclei that govern appetite and satiety. Activation suppresses glucagon, augments glucose-dependent insulin secretion, slows gastric emptying, and reduces caloric intake by altering central appetite signaling. The peripheral effects on glucose are real; the central effects on appetite are what move the scale.
Tirzepatide adds a second receptor. It binds GIPR — the receptor for glucose-dependent insulinotropic polypeptide — alongside GLP-1R, with relative affinity tuned closer to native GIP than to native GLP-1 [12]. The GIP arm of incretin biology was, until recently, the unfashionable one. GIP-receptor agonism in isolation doesn't lower glucose in type 2 diabetes, and GIP had a reputation as the incretin that did not work. The dual-agonist data forced a reread. In the presence of GLP-1R activation, GIPR activation appears to add an independent effect on adipose-tissue handling, central satiety signaling, and possibly nausea-pathway modulation in a way that allows higher effective dosing.
The honest read is that we don't yet know how much of tirzepatide's edge is GIPR-specific versus simply a function of being a more aggressively dosed, longer-titrated incretin agent. The pharmacology suggests both contribute. The clinical data show the gap is real regardless of which receptor takes the credit.
Liraglutide (Saxenda for obesity, Victoza for diabetes) is where this story begins clinically. The SCALE program, published across 2015–2016, established that a GLP-1 receptor agonist titrated to 3.0 mg daily could produce roughly 8% total body-weight loss at 56 weeks in adults with obesity [11]. At the time, that was the strongest evidence base any anti-obesity pharmacotherapy had ever produced. It's also, by current standards, modest.
Liraglutide is dosed daily, which is the first strike against it commercially and clinically. Discontinuation rates in SCALE ran high — both for GI intolerance and for the burden of seven injections a week. Eight percent weight loss is enough to produce meaningful metabolic improvement in many patients, but it sits below the threshold (roughly 10–15%) at which obesity-related comorbidities reliably remit.
Liraglutide is still on the market, still approved, and still occasionally the right answer — usually when a patient has tolerated it well in the past, when daily dosing is preferred for a specific reason, or when access to weekly agents is blocked. It's the floor against which the weekly agents are measured. The floor is now several rooms below where most prescribing happens.
The STEP program established semaglutide 2.4 mg weekly as the new standard for pharmacologic weight management. STEP 1, published in 2021, randomized 1,961 adults with obesity and no diabetes to semaglutide 2.4 mg or placebo for 68 weeks, both arms on lifestyle counseling. The semaglutide arm lost a mean of 14.9% of body weight versus 2.4% on placebo; roughly a third of treated patients lost 20% or more [3].
STEP 2 ran the same drug in adults with type 2 diabetes and reported smaller absolute losses — about 9.6% — a pattern that holds across the incretin class, in which diabetic populations consistently lose less weight than non-diabetic populations at equivalent doses [11]. STEP 5 extended treatment to 104 weeks and showed weight loss plateaued near 15% and was largely maintained with continued dosing.
The STEP data did two things. They established that a GLP-1 monoagonist, dosed weekly and titrated over four months, could produce weight loss in a range previously seen only with bariatric surgery's lower-tier procedures. They also established that the weight stays off only as long as the drug is taken — the withdrawal extension of STEP 1 showed regain of roughly two-thirds of lost weight within a year of stopping. Whatever the agent's mechanism, it's suppressive rather than curative.
SURMOUNT-1, published in NEJM in 2022, randomized 2,539 adults with obesity and no diabetes to tirzepatide 5, 10, or 15 mg weekly versus placebo for 72 weeks. Mean weight loss at the 15 mg dose was 20.9% [2]. Over half the patients on the top two doses lost at least 20% of body weight; roughly a third lost 25% or more. The placebo arm lost 3.1%.
The numbers were striking enough that the trial reframed expectations. Twenty percent total body-weight loss had been the rough threshold separating surgical from medical results. SURMOUNT-1 collapsed the distinction at the upper dose.
SURMOUNT-2 ran the same drug in type 2 diabetes and again showed the diabetes discount — roughly 13–16% weight loss across doses, lower than non-diabetic patients but still well above semaglutide's diabetic-population results. SURMOUNT-3 examined tirzepatide added after intensive lifestyle intervention and showed additional weight loss layered on top of the lifestyle response. SURMOUNT-4 addressed the maintenance question by withdrawing tirzepatide after a lead-in: patients continued on drug maintained and modestly extended their weight loss, while patients switched to placebo regained roughly fourteen percentage points over the next year [8]. The drug works while you take it. It stops working when you stop.
SURPASS-2, published in 2021, compared tirzepatide at 5, 10, and 15 mg weekly against semaglutide 1.0 mg weekly in 1,879 adults with type 2 diabetes inadequately controlled on metformin. At 40 weeks, tirzepatide at all three doses produced greater HbA1c reductions than semaglutide 1.0 mg, and greater weight loss: 7.6, 9.3, and 11.2 kg for tirzepatide 5/10/15 mg versus 5.7 kg for semaglutide [5].
Two limits on reading SURPASS-2 across the diabetes/obesity divide. The semaglutide comparator was dosed at 1.0 mg — the diabetes maintenance dose at the time of the trial — not at 2.4 mg, the obesity dose. The trial enrolled diabetic patients, whose weight-loss response to incretins is consistently attenuated. So while SURPASS-2 established tirzepatide's glycemic and weight superiority over a lower-dose semaglutide in diabetes, it did not — and was not designed to — answer the obesity question. That required SURMOUNT-5.
SURMOUNT-5 randomized 751 adults with obesity (BMI ≥30, or ≥27 with at least one comorbidity) and no diabetes to maximum tolerated doses of tirzepatide (10 or 15 mg weekly) or semaglutide (1.7 or 2.4 mg weekly) for 72 weeks. Both arms received the same lifestyle counseling. The primary endpoint was percent change in body weight at week 72 [1].
Tirzepatide produced a mean weight loss of 20.2%. Semaglutide produced 13.7%. The absolute difference was 6.5 percentage points; the relative difference was roughly 47% more weight loss on tirzepatide. Secondary endpoints — proportion of patients achieving ≥15%, ≥20%, and ≥25% weight loss; waist-circumference reduction; lipid and metabolic improvements — all favored tirzepatide, in most cases by margins that exceeded the absolute weight-loss gap [1].
What SURMOUNT-5 closed is the question of whether the cross-trial comparison was misleading. It was not. The tirzepatide advantage (6.5 points) is consistent with what indirect comparison of SURMOUNT-1 versus STEP 1 had suggested (roughly 6 points) [4]. The trial was a confirmation, run with the rigor needed to remove the cross-trial caveat that had let clinicians defer the comparison.
What SURMOUNT-5 doesn't tell us: whether the gap holds in subgroups (older adults, men, patients with cardiovascular disease), whether it translates into a corresponding advantage in cardiovascular event reduction (semaglutide has SELECT in obesity without diabetes; tirzepatide's SURPASS-CVOT tested a diabetes population against an active comparator), or whether a fraction of patients who tolerate semaglutide poorly would do better on tirzepatide and vice versa. The mean advantage is established. The individual prescribing decision still requires judgment.
The titration schedules differ in ways that matter clinically. Both drugs start low and climb to minimize GI side effects, both are subcutaneous weekly injections, both reach maintenance over months not weeks.
Semaglutide for obesity (Wegovy): 0.25 mg weekly for four weeks, then 0.5, 1.0, 1.7, and 2.4 mg, each held for four weeks. Total titration to maintenance is roughly sixteen to twenty weeks. If a patient stalls at an intermediate dose for tolerability, the label allows holding there longer or staying below 2.4 mg permanently.
Tirzepatide for obesity (Zepbound): 2.5 mg weekly for four weeks, then escalated by 2.5 mg every four weeks to a maintenance dose of 5, 10, or 15 mg. Titration to maximum is roughly twenty to twenty-four weeks. As with semaglutide, the label permits holding at a tolerated intermediate dose.
The titration arc is itself part of the side-effect story. The nausea, vomiting, and diarrhea that drive most discontinuations cluster around dose escalations — the week after each step up — and typically attenuate within one to two weeks at the new dose. Patients who quit usually quit during a step-up, not during steady-state dosing. Slowing the titration, holding at a lower dose, or returning to a previously tolerated dose are all reasonable when GI symptoms are limiting.
GI events dominate the tolerability profile of both drugs. In SURMOUNT-5, the most common adverse events on tirzepatide were nausea (44%), diarrhea (25%), vomiting (21%), and constipation (19%). On semaglutide, nausea ran 50%, diarrhea 25%, vomiting 24%, constipation 18% [1]. Read carefully: semaglutide produced numerically more nausea and vomiting than tirzepatide in this trial, despite tirzepatide producing more weight loss. The dual-agonist mechanism doesn't appear more emetogenic at the doses used; if anything the GIP arm may attenuate nausea relative to pure GLP-1 monoagonism, though the data on that are still preliminary.
Treatment discontinuation due to adverse events ran roughly 6% in both arms of SURMOUNT-5 — a number that surprises clinicians who expect higher dropout given symptom prevalence [1]. The interpretation: most patients have nausea, most tolerate it through dose escalation, and most stay on the drug. The ones who quit tend to quit early and decisively rather than slowly attrit.
Serious adverse events worth flagging across the class: acute pancreatitis (rare, labeled warning, mechanism uncertain), cholelithiasis (more common with rapid weight loss regardless of drug, plausibly potentiated by gallbladder hypomotility from GLP-1 effect), thyroid C-cell tumors (boxed warning based on rodent carcinogenicity data, contraindicated in patients with personal or family history of medullary thyroid carcinoma or MEN2, human relevance still uncertain after a decade of post-marketing surveillance), and the lean-mass-loss question now being investigated more rigorously.
The real-world cohort study by Rodriguez and colleagues, published in JAMA Internal Medicine in 2024, found broadly similar rates of clinically real GI adverse events between tirzepatide and semaglutide users in propensity-matched EHR data [6]. The network meta-analysis by Karagiannis and colleagues, published in Diabetologia in 2024, reached the same conclusion for diabetic populations [7]. The tolerability profiles are comparable. The efficacy profiles are not.
Both drugs carry weight-management indications keyed to the same BMI thresholds, inherited from the Saxenda labeling and now standard across the class.
Zepbound (tirzepatide for obesity): approved November 2023 for chronic weight management in adults with BMI ≥30, or ≥27 with at least one weight-related comorbidity (hypertension, dyslipidemia, type 2 diabetes, obstructive sleep apnea, cardiovascular disease). The label also includes a 2024 expansion for obstructive sleep apnea in adults with obesity.
Wegovy (semaglutide for obesity): approved June 2021 for chronic weight management in adults with the same BMI criteria. In 2023–2024 the label expanded to include cardiovascular risk reduction in adults with established cardiovascular disease and overweight or obesity, on the strength of the SELECT trial showing a 20% reduction in major adverse cardiovascular events [9].
Mounjaro and Ozempic remain approved for type 2 diabetes only, despite containing the same active molecules as Zepbound and Wegovy respectively. The bifurcated branding is commercial, not pharmacologic. Saxenda (liraglutide 3.0 mg daily) retains its 2014 obesity indication but is now rarely the first choice for new starts.
The compounded GLP-1 market that grew up around shortage-listed semaglutide and tirzepatide in 2022–2024 is largely closed. The FDA removed tirzepatide from the shortage list in late 2024 and semaglutide in early 2025. Grace periods for 503A and 503B compounding ran into mid-2025. In April 2026 the FDA proposed permanently excluding semaglutide, tirzepatide, and liraglutide from the 503B Bulks List, which would end the residual pathway by which outsourcing facilities had continued some mass production.
What remains: patient-specific compounding under 503A, where a licensed clinician documents that a specific patient can't use the commercially available product for a specific clinical reason. This is a narrow exception, not a market. Most patients who were on compounded GLP-1s through 2024 are now on branded product, off treatment, or in legal grey zones the FDA has been actively investigating.
The practical effect on prescribing is that the drug a patient gets is, again, the drug the manufacturer makes. The pricing conversation has moved from compounded-versus-branded to insurance-coverage-versus-cash, which is a different problem with different answers.
The clinical summary, as honestly as the evidence supports.
For a patient with obesity, no diabetes, no cardiovascular disease, and access to both drugs, tirzepatide is the better first choice. SURMOUNT-5 establishes a roughly 6.5-point absolute advantage in mean weight loss, and the higher proportions of patients reaching ≥20% and ≥25% weight loss matter clinically because that's the range where comorbidities reliably improve [1]. The tolerability profile is at worst equivalent and possibly slightly better than semaglutide.
For a patient with established cardiovascular disease, semaglutide currently has the better-supported indication. SELECT showed a 20% MACE reduction in adults with cardiovascular disease and obesity [9]. SURPASS-CVOT has since read out: tirzepatide was non-inferior to dulaglutide for major adverse cardiovascular events in adults with type 2 diabetes and established atherosclerotic disease [10]. That is a real result, but it is not the same claim SELECT supports — SURPASS-CVOT was an active-comparator trial in diabetes, measured against a drug already known to cut cardiovascular events, whereas SELECT was placebo-controlled in obesity without diabetes. For the obesity indication specifically, semaglutide still has the stronger cardiovascular evidence. If a patient on tirzepatide has good weight loss and stable cardiovascular disease, switching isn't warranted; if a patient is choosing between the two and CV risk is the dominant concern, semaglutide wins on labeled indication.
For a patient with type 2 diabetes and obesity, both drugs work, both are indicated, and the choice often comes down to glycemic target, insurance coverage, and tolerability. SURPASS-2 showed tirzepatide superior to semaglutide 1.0 mg for both HbA1c and weight [5]; whether it remains superior to semaglutide 2.0 mg (the higher diabetes dose now available) is suggested by network meta-analysis but not established head-to-head [7].
For a patient who has tolerated liraglutide well in the past, or who specifically prefers daily over weekly dosing, liraglutide remains an option. It's the option with the most modest efficacy and the highest injection burden. It's rarely the right answer for a new start.
For a patient who fails one agent on tolerability — meaning intolerable GI symptoms that don't resolve with slower titration or dose reduction — switching to the other class member is a reasonable next step. The cross-tolerability data are limited, but clinical experience suggests a meaningful fraction of patients who can't tolerate one agent can tolerate the other.
The question worth asking in clinic, after the prescribing decision is made: is the patient prepared to take this drug indefinitely. Both SURMOUNT-4 and the STEP withdrawal extensions show weight regain on discontinuation [8]. This is suppressive pharmacology. It works while it's being taken. The conversation about whether and how to stay on these drugs for years is the one prescribers should be having with patients before the first injection, not after the first plateau.
In the head-to-head SURMOUNT-5 trial, yes: 20.2% versus 13.7% mean body-weight loss at 72 weeks at maximum tolerated doses [1]. Individual response varies, and some patients tolerate semaglutide better.
Yes — switching to the other agent is a reasonable step for someone who can't tolerate one despite slower titration. Cross-tolerability data are limited but clinical experience supports it.
Semaglutide currently has the stronger labeled cardiovascular indication, based on the SELECT trial's ~20% reduction in major adverse cardiovascular events [9]. Tirzepatide's dedicated CV-outcomes trial, SURPASS-CVOT, reported in 2025: non-inferior to dulaglutide for major adverse cardiovascular events in adults with type 2 diabetes and atherosclerotic disease [10]. That is not the same as SELECT — an active comparator in diabetes rather than placebo in obesity — so for the obesity indication semaglutide remains the better-evidenced choice.
Largely, yes. Both the SURMOUNT-4 and STEP withdrawal data show substantial regain after discontinuation [8]. These drugs are suppressive, not curative, and are designed for long-term use.
Editorial note: Informational only — not medical advice. Decisions about GLP-1 receptor agonist therapy should be made with a licensed clinician familiar with your medical history, comorbidities, and current medications. See our methodology. § 15 / References 1. Aronne LJ, et al. Tirzepatide vs semaglutide for the treatment of obesity (SURMOUNT-5). N Engl J Med 2025. PMID 40353578. (Tier A — head-to-head RCT) https://pubmed.ncbi.nlm.nih.gov/40353578 2. Jastreboff AM, et al. Tirzepatide once weekly for obesity (SURMOUNT-1). N Engl J Med 2022. (Tier A — RCT) https://www.nejm.org/doi/full/10.1056/NEJMoa2206038 3. Wilding JPH, et al. Once-weekly semaglutide 2.4 mg in overweight or obesity (STEP 1). N Engl J Med 2021. (Tier A — RCT) https://www.nejm.org/doi/full/10.1056/NEJMoa2032183 4. Comparative meta-analysis of GLP-1 receptor agonists for weight loss (semaglutide vs liraglutide / dulaglutide / tirzepatide; provides cross-trial comparative context). Front Pharmacol 2025. (Tier A — meta-analysis) https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2025.1438318/full 5. Karagiannis T, et al. Tirzepatide vs semaglutide and other agents in type 2 diabetes (incl. SURPASS-2 head-to-head). Diabetologia 2022. PMID 35579691. (Tier A — meta-analysis) https://pubmed.ncbi.nlm.nih.gov/35579691/ 6. Rodriguez PJ, et al. Tirzepatide vs semaglutide for weight loss — real-world matched cohort. JAMA Intern Med 2024. PMID 38976257. (Tier A — cohort) https://pubmed.ncbi.nlm.nih.gov/38976257/ 7. Karagiannis T, et al. Tirzepatide vs semaglutide — network meta-analysis (T2D). Diabetologia 2024. PMID 38613667. (Tier A — network meta-analysis) https://pubmed.ncbi.nlm.nih.gov/38613667/ 8. Aronne LJ, et al. Continued tirzepatide vs withdrawal for maintenance of weight reduction (SURMOUNT-4). JAMA 2024. PMID 38078870. (Tier A — RCT) https://pubmed.ncbi.nlm.nih.gov/38078870/ 9. SELECT — semaglutide 2.4 mg and cardiovascular outcomes (NCT03574597; ~20% MACE reduction). (Tier A — RCT) https://clinicaltrials.gov/study/NCT03574597 10. Galli M, et al. Cardiovascular outcomes of GLP-1 receptor agonists — meta-analysis. J Am Coll Cardiol 2025. PMID 40892610. (Tier A — meta-analysis) https://pubmed.ncbi.nlm.nih.gov/40892610/ 11. Moiz A, et al. GLP-1 receptor agonists for weight loss in adults without diabetes — systematic review (incl. liraglutide/SCALE and semaglutide). Ann Intern Med 2025. PMID 39761578. (Tier A — systematic review) https://pubmed.ncbi.nlm.nih.gov/39761578/ 12. Nauck MA, D'Alessio DA. Tirzepatide — the GIP/GLP-1 dual mechanism and clinical context. Cardiovasc Diabetol 2022. PMID 36050763. (Tier B — peer-reviewed review) https://pubmed.ncbi.nlm.nih.gov/36050763/ --- Editorial note: Informational only — not medical advice. Head-to-head figures drawn from separate trials are not equivalent to a direct comparison, and individual response varies. Which of these is appropriate — if either is — is a decision for you and a licensed healthcare provider familiar with your medical history. See our methodology. Last reviewed July 2026.