What is Mazdutide? What it does, benefits and what the evidence shows

Mazdutide is a synthetic peptide and dual agonist of the glucagon (GCG) and GLP-1 receptors, structurally related to the gut hormone oxyntomodulin. Innovent Biologics develops it in China under an exclusive license agreement with Eli Lilly and Company for its development and commercialization there. China's National Medical Products Administration (NMPA) approved it for chronic weight management in adults with obesity or overweight in June 2025, and for glycemic control in adults with type 2 diabetes in September 2025. It has not been approved by FDA, EMA, or any other major regulator, and it is not sold under a US brand name.

What mazdutide is and how it works

Innovent describes mazdutide as the world's first dual GCG/GLP-1 receptor agonist approved for either weight loss or glycemic control, built on a mammalian oxyntomodulin scaffold (Innovent, PRNewswire, June 27, 2025). Oxyntomodulin is a natural gut hormone that activates both the glucagon receptor and the GLP-1 receptor; mazdutide is a modified, longer-acting analogue of it rather than a copy of semaglutide or tirzepatide, which target GLP-1 alone or GLP-1 plus GIP.

A 2026 pharmacology review describes the two receptors doing different jobs. GLP-1 receptor activation works the way it does in other incretin drugs — suppressing appetite and slowing gastric emptying to cut energy intake. Glucagon receptor activation is the differentiating piece: in the liver, it "stimulates hepatic cAMP-PKA signaling, which enhances fatty acid oxidation and mitochondrial thermogenesis," which the authors describe as a "dual metabolic engine" that raises energy expenditure on top of appetite suppression (Deng et al., World Journal of Pharmacology, 2026). The same review notes mazdutide is "structurally derived from the GLP-1 scaffold and modified by lipidation to extend its plasma half-life," which is what allows infrequent injection rather than a short-acting peptide requiring frequent dosing.

Innovent holds the commercial rights in China under its license from Lilly, and Lilly itself has run a US Phase 2 trial of mazdutide; retatrutide and tirzepatide are separate, differently engineered compounds, not mazdutide under another name.

That distinction matters for anyone comparing grey-market listings. A vial labeled "mazdutide" is not a Chinese version of retatrutide or tirzepatide — it's a different sequence, developed and reviewed on a separate clinical program, with its own approval history and its own evidence base. Marketing copy that blurs the three together, or implies mazdutide is simply "tirzepatide from China," isn't describing the actual regulatory or scientific record.

Where mazdutide stands, regulator by regulator

It's worth being precise about what "approved" means here, because this is the point readers most often get wrong. China's NMPA is a national medicines regulator with its own review standards, distinct from FDA or the EMA; an NMPA approval means the agency reviewed Innovent's trial data and cleared the drug for prescription use inside China, under a Chinese label, for Chinese patients. It does not mean FDA has reviewed the same data, and it does not create any prescribing pathway in the US, EU, or anywhere outside China. We found no FDA submission, no EMA filing, and no other national regulator's approval for mazdutide as of this writing — only the two NMPA approvals described above. Anyone selling or using mazdutide outside a Chinese prescription is doing so entirely outside any regulator's oversight, which is the market this article's lab-testing section describes.

What people use it for

What the research actually shows

Weight loss — GLORY-1, the pivotal Chinese trial

GLORY-1 randomized 610 Chinese adults with obesity (BMI of at least 28) or overweight with a weight-related condition (BMI 24 to under 28) to one of two mazdutide doses or placebo for close to a year, with body weight change as a co-primary endpoint. At week 48, the two mazdutide dose groups lost an average of 12.05% and 14.84% of body weight versus 0.47% on placebo (Innovent's efficacy estimand); in the NEJM paper's treatment-policy analysis the figures were 11.00% and 14.01% versus a 0.30% gain on placebo. About half of people on the higher dose lost at least 15% of their body weight (50.6% in Innovent's release, 49.5% in the NEJM paper), versus about 2% on placebo (Innovent, reporting the GLORY-1 results published in the New England Journal of Medicine). The same release reports average heart rate increased modestly (about 2.6 beats per minute) in both mazdutide groups by the end of the trial, and states no signal of increased cardiovascular risk was observed. This is the trial NMPA approval for weight management was based on; it has not been replicated or reviewed by FDA.

Glycemic control — DREAMS-1 and DREAMS-2

Innovent's own results summary for the type 2 diabetes indication cites two registered trials. In DREAMS-1, a monotherapy study, HbA1c fell further on mazdutide than on placebo, and the majority of people on the higher studied dose reached an HbA1c under 7% by the trial's efficacy readout; nearly two-thirds hit that HbA1c target while also losing at least 5% of body weight. In DREAMS-2, a combination-therapy study measured against another approved diabetes drug (dulaglutide) rather than placebo, both mazdutide arms produced substantially larger weight loss than the comparator by the end of the trial (Innovent NMPA approval announcement, September 19, 2025). Both trials were conducted in Chinese adults with type 2 diabetes and have since been published in Nature (2026), and the only controlled data we found from outside China is the US Phase 2 obesity trial described above.

What's not established

The main efficacy data still come from China, although a US-based Phase 2 trial of 179 adults with obesity was published in The Lancet Diabetes & Endocrinology in August 2026 (weight change at 32 weeks of -7.3% to -18.1% across mazdutide dose groups versus -0.9% on placebo). We found no FDA or EMA review of any kind, and no long-term (multi-year) outcome data — the pivotal trials run under a year. Claims that circulate online treating mazdutide as interchangeable with tirzepatide or retatrutide aren't supported by anything we could verify: it's a different molecule with a different receptor pair (glucagon plus GLP-1, rather than GIP plus GLP-1), and cross-compound efficacy hasn't been studied head-to-head in a trial we could find.

Side effects and risks

In the trials Innovent has reported, the most common adverse events were nausea, diarrhea and vomiting, described as mostly mild to moderate, consistent with the GI-related side effects typical of incretin drugs as a class (Innovent, GLORY-1 results). A 2026 pharmacology review comparing dual agonists like mazdutide with single-target GLP-1 drugs concludes that GLP-1 receptor agonists generally "exhibit significantly fewer side effects" than newer multi-receptor agonists, and that mazdutide's adverse-event profile so far has been dominated by the same GI symptoms (Deng et al., World Journal of Pharmacology, 2026).

Because mazdutide has only been reviewed by one regulator, mostly on trials run in one country, there's no FDA label, no long-term safety monitoring outside those trials, and no independent post-market pharmacovigilance data we could find covering years of use. Anyone taking it outside a Chinese prescription — which describes essentially every grey-market buyer — is taking it with less real-world safety information behind it than an FDA-approved GLP-1 drug would have, on top of the manufacturing-quality risk that applies to any unregulated peptide vial.

There's also a practical safety gap that has nothing to do with the trial data: an NMPA-approved drug is dispensed under a Chinese physician's supervision, with baseline screening and follow-up built into the prescribing relationship. None of that infrastructure exists for a vial bought on the grey market. The trial safety signals above — GI symptoms, a small heart-rate increase, no observed cardiovascular signal over the trial period — describe people who were screened, monitored, and followed by a clinical trial team. They don't describe what happens when the same molecule is used without any of that oversight, and we found no published data that does describe that scenario.

On that manufacturing-quality risk specifically: our data on mazdutide vials sold on the grey market is thin (see below), but as of September 29, 2026, purity and dose-versus-label numbers so far look better than several other compounds we track — this compound's quality problem, if there is one, hasn't shown up yet in a small sample.

What the lab tests show

As of September 30, 2026, our database includes 7 mazdutide vials from 4 vendors, drawn from 4 lab reports dated August 2024 through May 2026. 7 of the 7 vials had a purity result; 6 came in at 99% or higher, with a median purity of 99.53%. Dose-versus-label ran a median of 4.5% over label, with 4 of the vials within 5% of label (6 were within 10% as of September 29, 2026). As of September 29, 2026, no vials had failed our definition (empty, under 95% pure, more than 20% under label, or more than 50% over) in this sample.

As of September 29, 2026, that's a very small sample — 4 vendors is not enough to draw a real conclusion about the compound as a category, and a single new report could shift these numbers substantially. Here's how the ranked vendors for mazdutide look right now; the table shows current verdicts and flags any vendor whose failed-vial rate is above 5%, though with only 7 vials tested for this compound as of September 29, 2026, don't expect it to be long:

Buying smarter

If you're evaluating a specific vendor's mazdutide, check their full record on our mazdutide results page rather than relying on a handful of reports — a vendor needs five or more lab reports before we rank them, and with only about 4 reports total for this compound across the whole database as of September 29, 2026, most vendors won't clear that bar yet. Whatever COA a vendor posts is a starting point, not proof: verify it against how to read a peptide COA and check for the patterns in fake peptide COAs before trusting a number you didn't see tested yourself.

Given how thin the data is for this specific compound, testing your own vial adds more information than it would for a heavily tested peptide like retatrutide or tirzepatide, where hundreds of data points already exist. How to get a peptide tested walks through ordering independent identity, purity and quantity testing.

If your own report comes back with a fail, stop using that batch, tell anyone else who bought from the same run, bring the report to the vendor, and submit it to VendorVerify — for a compound with this few reports, one submission moves the needle more than it would almost anywhere else in our database.

Database figures without a date update automatically from VendorVerify's lab results; dated figures are snapshots. This isn't medical advice; mazdutide is sold for research use only in most markets and has not been evaluated by FDA for safety or effectiveness.