“Alternative” usually means one of three different things, and the useful answer depends entirely on which one is being asked.
A cheaper version of the same drug. A different drug because this one is intolerable. Or something else entirely because a GLP-1 is not available at all.
The ladder
| Option | Route | Approx. average weight loss | Notes |
|---|---|---|---|
| Bariatric surgery | Procedure | ~25–30% | One-time; irreversible |
| Tirzepatide (Zepbound) | Weekly injection | ~21% | Largest of any approved drug |
| Semaglutide (Wegovy) | Weekly injection | ~15% | Also licensed for CV risk |
| Phentermine–topiramate (Qsymia) | Daily oral | ~9–10% | Most effective non-GLP-1 |
| Naltrexone–bupropion (Contrave) | Daily oral | ~5% | Helps some with food cravings |
| Orlistat (Xenical, Alli) | Oral with meals | ~3% | Blocks fat absorption |
| Phentermine alone | Daily oral | ~5% short-term | Licensed for short-term use only |
| Metformin | Daily oral | ~2–3% | Off-label for weight; very cheap |
Trial populations and durations differ, so these are not perfectly comparable — but the ordering is robust and the gaps are wide enough that small methodological differences do not reorder it.
~9–10%
Average weight loss with phentermine-topiramate — the best non-GLP-1 option
EQUIP, Obesity 2012
~5%
Average weight loss with naltrexone-bupropion
COR-I, Lancet 2010
~3%
Average weight loss with orlistat, the oldest option still in wide use
Staying within the class
The other molecule. Tolerance is molecule-specific, so someone who cannot get past 0.5 mg of semaglutide sometimes does fine on tirzepatide, and occasionally the reverse. You restart at the bottom of the new ladder either way — the mechanics are in switching between GLP-1s, and the efficacy comparison is in semaglutide versus tirzepatide.
The other brand of the same molecule. Ozempic and Wegovy are both semaglutide; Mounjaro and Zepbound are both tirzepatide. Which one you are prescribed is largely an indication and coverage question — see Ozempic versus Wegovy.
Oral semaglutide. Rybelsus at diabetes doses, and a higher-dose oral formulation for obesity — see the obesity pill.
What is coming. Orforglipron is a non-peptide oral GLP-1 with the enormous manufacturing advantage of not being a peptide; retatrutide and CagriSema are targeting effect sizes above tirzepatide. See orforglipron and the next wave.
The non-GLP-1 drugs, honestly
Phentermine–topiramate (Qsymia). The most effective thing here that is not an injection. A stimulant appetite suppressant combined with an anticonvulsant. Real contraindications: it is teratogenic, so pregnancy must be excluded and contraception used; it raises heart rate; and topiramate causes cognitive fog and paraesthesia in a meaningful minority.
Naltrexone–bupropion (Contrave). Modest average effect, but it works on reward and craving rather than on satiety, which suits some people well — particularly those who describe the problem as food noise rather than portion size. Contraindicated with opioids and in uncontrolled hypertension, and it lowers seizure threshold.
Phentermine alone. Cheap, generic, effective in the short term, and licensed only for short-term use in most jurisdictions. Widely prescribed for longer than that. A stimulant, with the blood pressure and sleep consequences that implies.
Orlistat. Blocks about a third of dietary fat absorption. The effect is modest and the side effects are as socially unforgiving as they are famous. Also reduces absorption of fat-soluble vitamins.
Metformin. Not a weight-loss drug. Produces two to three kilograms in most people, is extremely cheap and extremely well characterised, and is genuinely useful in insulin-resistant conditions — see GLP-1s and PCOS.
The complete class directory, including the older agents a formulary may route you through, is in every GLP-1 medication.
The routes that are not different drugs
Compounded semaglutide or tirzepatide. Same molecules, different supply chain, different risk profile. Worth reading choosing a compounding pharmacy and shortages and compounding before deciding.
Buying abroad. Genuinely cheaper in most countries, with real complications — see generic semaglutide outside the US.
Surgery. The most effective option available, and the comparison deserves its own treatment: GLP-1s versus bariatric surgery.
Supplements. Not an alternative in any meaningful sense — see the “natural Ozempic” question.
Step therapy will often make you fail a three percent drug before it lets you try a fifteen percent one. That is a budgeting decision wearing a clinical costume.
The step therapy problem
Worth understanding, because it explains a lot of otherwise baffling insurance behaviour.
Many plans require you to try and fail an older, cheaper agent before approving a GLP-1. On paper that is prudent sequencing. In practice it means spending three to six months on a drug with a third of the effect, in order to document that it did not work.
If you are in that position, the documentation matters more than the outcome — see appealing a prior authorisation, and check whether you meet an indication that routes around the weight-loss exclusion entirely, such as cardiovascular risk or obstructive sleep apnoea.
The realistic framing
If you can access a GLP-1, it is the better drug, and by a wide margin.
If you cannot, Qsymia at nine or ten percent is a genuine treatment rather than a consolation prize, and a great many people would take that result happily. What is not worth doing is spending a year on something at three percent while believing you are treating the problem — that is the version of this that wastes the most time.