Metformin vs Semaglutide: Two Very Different Drugs for Weight

These two get compared constantly and they are not really competitors. One is a sixty-year-old oral generic that costs a few dollars a month and produces around 2 percent weight loss as a side effect of doing something else. The other is a weekly injection that costs hundreds and produces around 15 percent because that is what it was designed to do.

The comparison is still worth making, because the cheap one is frequently the right answer and the expensive one is frequently prescribed where the cheap one was never tried. This page puts the actual numbers side by side and is explicit about which question each drug answers. For the mechanism in detail see how semaglutide works. This is research use information and none of it is medical advice.

The Numbers

A seven-fold difference in effect for something like fifty times the price. Both halves of that sentence are the point.

What Metformin Actually Does for Weight

Metformin is a biguanide, first-line for type 2 diabetes, and it works mainly by reducing how much glucose the liver produces and improving how well tissues respond to insulin. Weight loss is not its purpose and it is not approved for it anywhere; it is a modest and welcome side effect.

The Diabetes Prevention Program is the best long-term evidence. Participants on metformin lost an average of 2.1kg over the 2.8 year study, and about 2.5kg across the combined DPP and follow-up period, against roughly 0.2 percent in the placebo arm. That is small, and two things make it more interesting than the headline.

It is durable. Weight loss on metformin held for at least ten years of continued treatment, which is unusual, and stands in contrast to what SURMOUNT-4 showed happens when a GLP-1 is stopped. And the average conceals a split: roughly 30 percent of participants lost more than 5 percent of body weight in the first year, and that responder group averaged 6.2 percent below baseline after fifteen years. Metformin has a minority of genuine responders hidden inside an unimpressive mean.

Why Semaglutide Does So Much More

Because it targets appetite directly rather than incidentally. Semaglutide is a long-acting GLP-1 receptor agonist: it slows gastric emptying so a smaller meal produces the fullness a larger one used to, and it acts on hypothalamic and brainstem circuits that set the background drive to eat. The result in STEP 1 was a 14.9 percent mean reduction over 68 weeks against 2.4 percent on placebo.

Metformin does not have a comparable appetite mechanism. Whatever weight effect it has comes downstream of its metabolic actions, which is exactly why it is measured in kilograms while semaglutide is measured in percentages of body weight.

Cost Is the Real Comparison

Generic metformin costs single-digit dollars a month almost anywhere. Semaglutide without insurance coverage runs into the hundreds, and the routes and their tradeoffs are laid out in semaglutide cost without insurance, with dose arithmetic in the semaglutide dosage calculator.

Put against the effect sizes, metformin delivers roughly 2 percent for almost nothing and semaglutide roughly 15 percent for a real monthly commitment. Whether that is good value depends entirely on what you need. For someone with 5kg to lose and insulin resistance in the background, metformin may be the sensible first move. For someone with 30kg to lose and a weight-related complication, metformin is not going to get there and the arithmetic changes.

Side Effects, Compared

Both are gastrointestinal drugs in terms of what people notice, and the similarity ends there.

Metformin: diarrhoea and nausea, common early and usually improving, and substantially reduced by extended-release formulations. Long-term use is associated with reduced B12 absorption, which is worth monitoring on a multi-year course. Lactic acidosis is the serious one and is rare, mainly relevant in significant kidney impairment.

Semaglutide: nausea, diarrhoea, vomiting and constipation, clustered around dose escalation. Beyond those, a warning for pancreatitis, an increased rate of gallbladder disease, and a boxed warning for thyroid C-cell tumours based on rodent studies, with a contraindication in anyone with a personal or family history of medullary thyroid carcinoma or MEN 2. Detail is in semaglutide side effects.

That is a genuinely different risk profile, not a matter of degree, and it is part of why a cheap drug with a sixty-year safety record is not simply the weaker option.

They Are Often Used Together

The framing as a choice is partly false. In type 2 diabetes, metformin and a GLP-1 receptor agonist are routinely prescribed together, because the mechanisms do not overlap: one reduces hepatic glucose output and improves insulin sensitivity, the other reduces intake and slows gastric emptying.

Neither causes hypoglycaemia on its own, which makes the combination more tolerable than adding either to insulin or a sulfonylurea. Whether adding metformin to semaglutide produces meaningfully more weight loss in people without diabetes has not been established in a dedicated trial, so that specific claim should be treated as plausible rather than demonstrated.

Which Question Are You Asking?

• Cheapest thing that might help a little? Metformin. Roughly 2 percent on average, durable, single-digit dollars, and a minority respond considerably better.
• Largest effect available? Semaglutide, or tirzepatide, which beat it head to head at 20.2 percent against 13.7 percent in SURMOUNT-5. See semaglutide vs tirzepatide.
• Insulin resistance or prediabetes as the main concern? Metformin has the prevention evidence, which is what the DPP was built to test.
• Oral only? Metformin, or oral semaglutide, covered in oral semaglutide.
• Worried about what happens on stopping? A real point in metformin's favour. Its effect held for a decade of continued use, while SURMOUNT-4 showed about 14 percent regain within a year of stopping a GLP-1.

Key Takeaways

• Metformin: about 2.1kg, roughly 2 percent, in the Diabetes Prevention Program. Semaglutide 2.4mg: 14.9 percent in STEP 1.
• Metformin is not approved for weight loss anywhere and is used off-label for it.
• Its effect is durable for at least ten years of continued treatment, and about 30 percent of people respond considerably better than the mean.
• Semaglutide targets appetite directly, which is why it is measured in percentages rather than kilograms.
• The cost gap is roughly fifty-fold and the effect gap roughly seven-fold.
• The risk profiles differ in kind, not only degree: semaglutide carries a boxed warning metformin does not.
• They are routinely combined in type 2 diabetes, and the mechanisms do not overlap.

Frequently Asked Questions

Is semaglutide better than metformin for weight loss?

For weight loss specifically, by a wide margin. Semaglutide 2.4mg produced a 14.9 percent mean reduction over 68 weeks in STEP 1, while metformin produced about 2.1kg, roughly 2 percent, in the Diabetes Prevention Program. Metformin is not approved for weight loss anywhere and its weight effect is incidental to its metabolic actions. Better for weight is not the same as better overall: metformin costs single-digit dollars, has a sixty-year safety record and does not carry a boxed warning.

How much weight can you lose on metformin?

About 2.1kg on average over the 2.8 years of the Diabetes Prevention Program, and roughly 2.5kg across the longer follow-up. The average hides a split worth knowing about: roughly 30 percent of participants lost more than 5 percent of body weight in the first year, and that responder group averaged 6.2 percent below baseline after fifteen years. So a minority do considerably better than the mean suggests.

Can you take metformin and semaglutide together?

Yes, and in type 2 diabetes the combination is routine. The mechanisms do not overlap: metformin reduces hepatic glucose production and improves insulin sensitivity, while semaglutide slows gastric emptying and reduces appetite. Neither causes hypoglycaemia on its own, which makes the pairing more tolerable than adding either to insulin or a sulfonylurea. Whether the combination produces meaningfully more weight loss in people without diabetes has not been tested in a dedicated trial.

Is metformin cheaper than semaglutide?

Dramatically. Generic metformin costs single-digit dollars a month in most places, while semaglutide without insurance coverage runs into the hundreds. Against effect sizes of roughly 2 percent and roughly 15 percent, that is about a fifty-fold cost difference for about a seven-fold difference in weight loss.

Which has worse side effects, metformin or semaglutide?

They differ in kind rather than only in degree. Metformin commonly causes diarrhoea and nausea early on, usually improving and much reduced by extended-release formulations, with reduced B12 absorption on long-term use and rare lactic acidosis mainly relevant in significant kidney impairment. Semaglutide causes gastrointestinal effects clustered around dose escalation, plus warnings for pancreatitis and gallbladder disease and a boxed warning for thyroid C-cell tumours with a contraindication in medullary thyroid carcinoma or MEN 2.

Does weight come back after stopping metformin?

This is one of the clearer points in metformin's favour. Weight loss on metformin was maintained for at least ten years of continued treatment in the Diabetes Prevention Program follow-up, and the effect is tied to adherence rather than producing dramatic rebound. By contrast SURMOUNT-4 showed about 14 percent weight regain within a year of stopping a GLP-1, because that class substitutes for an appetite signal rather than resetting one.