Metformin and Ozempic are both type 2 diabetes medicines, and on their own labels' placebo trials they lower blood sugar by a similar amount — they differ on weight, on side effects, and on what else they are approved to do. Metformin's label trial showed HbA1c down 1.4 points and no weight loss beyond placebo. Ozempic's showed HbA1c down 1.4 to 1.6 points and about 2.6 to 3.5 kg more weight loss than placebo, and Ozempic is approved for heart and kidney outcomes too. Neither is approved for weight loss. Most people asking this are not choosing one over the other: most of Ozempic's diabetes trials were run in people already taking metformin.
This page sets the two labels side by side, read on 28 October 2026, with the American Diabetes Association's 2026 Standards of Care for context. It does not say which medicine anyone should take. Metformin's own weight record, including the Diabetes Prevention Program, is on metformin for weight loss; Ozempic's full trial record is on Ozempic for type 2 diabetes.
The two labels side by side
| Metformin (generic, immediate release) | Ozempic (semaglutide injection) | |
|---|---|---|
| Class | biguanide | GLP-1 receptor agonist |
| How taken | tablet, usually once or twice daily | injection once weekly |
| US indications | blood-sugar control in type 2 diabetes | blood-sugar control in type 2 diabetes; reducing major cardiovascular events in type 2 diabetes with heart disease; reducing kidney decline, kidney failure and cardiovascular death in type 2 diabetes with chronic kidney disease |
| Weight loss an indication? | no | no (semaglutide for weight is Wegovy) |
| Boxed warning | lactic acidosis | thyroid C-cell tumours in rodents; human relevance not determined |
| Generic? | yes, many makers | no |
Sources: metformin hydrochloride tablets label (A-S Medication Solutions listing, effective 21 August 2024) and Ozempic label (effective 30 July 2026), both from the FDA's openFDA label records.
The placebo trials on each label
Each label carries one placebo-controlled trial of the medicine on its own in people with type 2 diabetes. They are the fairest like-for-like figures available, with one large caveat: they were run in different decades, in different people, and no trial compares metformin with Ozempic directly.
| Metformin label trial | Ozempic label trial (SUSTAIN 1) | |
|---|---|---|
| Length | 29 weeks | 30 weeks |
| People | 286, diet-treated, fasting glucose ~240 mg/dL | 388, diet and exercise only, mean BMI 33 |
| Starting HbA1c | 8.4% (placebo 8.2%) | 8.0–8.1% |
| HbA1c change | −1.4 vs placebo +0.4 | −1.4 (0.5 mg) and −1.6 (1 mg) vs placebo −0.1 |
| Starting weight | 201 lb (placebo 206 lb) | 89–97 kg |
| Weight change | −1.4 lb vs placebo −2.4 lb | −3.8 kg and −4.7 kg vs placebo −1.2 kg |
| Weight change in pounds | −1.4 vs −2.4 | about −8.4 and −10.4 vs −2.6 (our conversion) |
On the label's own trial, people on metformin lost slightly less weight than people on placebo. That is a single 29-week study, and other metformin studies show a modest loss — the 2026 ADA Standards grade its weight effect "neutral (potential for modest loss)". Ozempic's difference from placebo was 2.6 kg on 0.5 mg and 3.5 kg on 1 mg. At the higher weight-loss dose, semaglutide's results in type 2 diabetes are larger still: 9.6% against 3.4% at 68 weeks in STEP 2, the trial behind Wegovy's label.
One more number from metformin's label is worth having: in a second 29-week trial it ran against glyburide, weight on metformin alone fell 8.4 lb, against 0.7 lb on glyburide and a 0.9 lb rise on the combination. Metformin looks better on weight when the comparator is a drug that adds weight — the same pattern seen with Jardiance against glimepiride.
Side effects, from each label
| Adverse reaction | Metformin label trial (vs placebo) | Ozempic pooled placebo trials, 0.5 mg / 1 mg (vs placebo) |
|---|---|---|
| Diarrhoea | 53% (12%) | 8.5% / 8.8% (1.9%) |
| Nausea / vomiting | 26% combined (8%) | nausea 15.8% / 20.3% (6.1%); vomiting 5.0% / 9.2% (2.3%) |
| Flatulence | 12% (6%) | — |
| Abdominal pain or discomfort | 6% (5%) | 7.3% / 5.7% (4.6%) |
| Constipation | — | 5.0% / 3.1% (1.5%) |
| Stopped because of the side effect | diarrhoea: 6% | — (not in this table on the label) |
The two drugs both act on the gut but not in the same way: metformin's label trial is dominated by diarrhoea; Ozempic's by nausea. The ADA Standards note that metformin's gastrointestinal effects "can be mitigated by gradual dose titration and/or using extended-release formulation", and that metformin is associated with vitamin B12 deficiency, recommending periodic testing. Ozempic's side effects in more detail are on Ozempic side effects.
Where the American Diabetes Association places each
The ADA's Standards of Care in Diabetes—2026 (section 9) do not set a single first drug. They say choice should follow "a person-centered shared decision-making approach" weighing cardiovascular, kidney and weight effects, hypoglycaemia, cost and preference (recommendation 9.5). In the text:
- Metformin "historically has been the first-line treatment for type 2 diabetes" and "is effective and safe, is inexpensive and widely available, and reduces risks of microvascular complications, cardiovascular events, and death." It should not be started below an eGFR of 45 and is stopped below 30.
- GLP-1 receptor agonists — with SGLT2 inhibitors — should be part of the plan for people with established or high risk of atherosclerotic heart disease "irrespective of A1C" (9.7), and a GLP-1 receptor agonist is preferred in advanced kidney disease (9.11).
- Weight, in Table 9.2: metformin "neutral (potential for modest loss)"; GLP-1 receptor agonists "loss (intermediate to very high)".
These are the ADA's positions, reported here; they are applied to a particular person by their clinician.
Taken together
The two are more often combined than compared. Ozempic's label trials were largely run in people already on metformin: SUSTAIN 2 (56 weeks) added Ozempic or sitagliptin to "metformin and/or thiazolidinediones", and the trial of the 2 mg dose added it to metformin with or without a sulfonylurea. The ADA Standards also endorse starting with a combination in some people "to shorten time to attainment of individualized glycemic goals" (9.6). Whether one, the other, or both fits a given person is a question for their prescriber.
Cost
Metformin is an old generic sold by many makers; we did not read a published cash price for this page, and pharmacy prices vary. Ozempic's self-pay prices from Novo Nordisk, re-read on 26 October 2026, are $199 a month for a new patient's first two fills at 0.25 or 0.5 mg, then $349, and $499 at 2 mg; with commercial insurance that covers it, Novo's savings card applies. The detail is on Ozempic cost and the Ozempic savings card.
What this page cannot tell you
- A head-to-head result. None exists on either label; the comparison above is between two separate placebo trials.
- What happens in people without diabetes. Neither label covers them; semaglutide for weight in people without diabetes is Wegovy, and metformin's evidence there is on our metformin page.
- Which is right for anyone. That depends on kidney function, heart history, other medicines and cost, and is decided with a prescriber.
Sources
- Metformin hydrochloride tablets, US prescribing information (A-S Medication Solutions listing), effective 21 August 2024, set ID 011de1a5-1ac0-4831-9e8d-26ec79ba2205, openFDA label record, read 28 October 2026 — boxed warning, Table 1, section 14.
- Ozempic (semaglutide) injection, US prescribing information, Novo Nordisk, effective 30 July 2026, set ID 42bdd912-2393-44c4-b7e0-47672ca28991, openFDA label record, read 28 October 2026 — indications, Table 1, section 14.
- American Diabetes Association. "9. Pharmacologic Approaches to Glycemic Treatment: Standards of Care in Diabetes—2026." Diabetes Care 2026;49(Suppl 1), doi:10.2337/dc26-S009, PMID 41358900, read 28 October 2026.
- Davies M et al. "Semaglutide 2·4 mg once a week in adults with overweight or obesity, and type 2 diabetes (STEP 2)." Lancet 2021, PMID 33667417.
- Ozempic self-pay prices: our GLP-1 cost page, NovoCare re-read 26 October 2026.