Mazdutide is a weekly injection acting on both the glucagon and GLP-1 receptors, approved in China in 2025 and sold there by Innovent Biologics, which calls it the world's first of its class to be approved. In the US it is unapproved and unfiled; Eli Lilly's registered trials of it are phase 1 and 2 — but they test doses far above China's. This page reads what is published, registered and announced, as of 5 October 2026.
The closest Western equivalent, Boehringer Ingelheim's survodutide, works the same two receptors and is not approved. Why glucagon is added to a GLP-1 drug at all — liver fat and energy expenditure, at the cost of heart rate — is discussed on survodutide vs retatrutide, and glucagon itself has an entry on Peptide Lexicon's glucagon page.
Two owners, two programmes
Innovent describes mazdutide as "a mammalian oxyntomodulin (OXM) analogue" under "an exclusive license agreement with Eli Lilly and Company … in China". The two names in the literature are the two programmes: IBI362 is Innovent's, LY3305677 is Lilly's.
| Innovent (China) | Eli Lilly (outside China) | |
|---|---|---|
| Registered studies (5 Oct 2026) | 19 | 7 |
| Furthest stage | approved, two indications | phase 2 (registered trials) |
| Doses tested | 2–9 mg | up to 16 mg |
| Results posted on ClinicalTrials.gov | 0 | 1 (phase 2, NCT06124807) |
ClinicalTrials.gov API v2, 5 October 2026: 43 studies list mazdutide as an intervention; the other 17 are mostly investigator-run studies at Chinese hospitals, plus two other companies' trials that list it among their interventions.
China's approvals
| Date | Event | Source |
|---|---|---|
| 27 June 2025 | Approved for chronic weight management, adults with BMI ≥28, or ≥24 with at least one weight-related condition | Innovent announcement |
| 19 September 2025 | Approved for blood-sugar control in adults with type 2 diabetes | Innovent announcement |
| 25 November 2025 | Application for a 9 mg dose in moderate-to-severe obesity accepted for review | Innovent announcement |
Innovent's announcements call it "the world's first dual GCG/GLP-1 receptor agonist approved" for each use; neither it nor survodutide had a US approval on 5 October 2026. Whether the 9 mg application has been decided was not stated in any Innovent announcement we opened.
What the trials found
| Trial | People | Arms, length | Weight change | Source |
|---|---|---|---|---|
| GLORY-1 | 610 Chinese adults, BMI ≥28 or ≥24 + condition | 4 mg / 6 mg / placebo, 48 wk | −11.0% / −14.0% / +0.3% | NEJM 2025;392:2215 |
| GLORY-2 | 462 Chinese adults, BMI ≥30 | 9 mg / placebo, 60 wk | −16.65% / −1.50% | JAMA 2026;336:377 |
| DREAMS-1 | 320 adults, new type 2 diabetes | 4 mg / 6 mg / placebo, 24 wk | −5.61% / −7.81% / −1.26%; HbA1c −1.57 / −2.15 / −0.14 | Nature 2026;652:174 |
| DREAMS-2 | 731 adults, type 2 diabetes on oral drugs | 4 mg / 6 mg vs dulaglutide 1.5 mg, 28 wk | −3.78 and −5.76 points more than dulaglutide; HbA1c −0.24 and −0.30 more | Nature 2026;652:181 |
| DREAMS-3 | Chinese adults, type 2 diabetes + obesity | mazdutide vs semaglutide, 32 wk | −10.29% vs −6.00%; HbA1c −2.03 vs −1.84 | Innovent, 27 Oct 2025 (not yet peer-reviewed) |
| Lilly phase 2 | 179 US adults, obesity/overweight, no diabetes | 3–6, 10, 16 mg / placebo, 48 wk | at 32 wk: −7.3% / −15.6% / −18.1% / −0.9% | Lancet Diabetes Endocrinol, 21 Aug 2026 |
Every published trial above was funded by Innovent or Lilly, and company employees are among the authors of each paper.
GLORY-2 comes in two sizes
Innovent's announcement of GLORY-2 gave 18.55% against 3.02% at 60 weeks. The JAMA paper reports −16.65% against −1.50%, analysed in all 461 participants who received at least one dose. The announcement did not say which estimand or population its figure used, and both arms differ by one and a half to two points. The placebo-adjusted difference is close in both: 15.5 points in the announcement and 15.15 in the paper (our arithmetic for the first). Quote the journal figure, and the difference is the safer number to carry between them. One more registry detail: the GLORY-2 record (NCT06164873) still showed "Unknown" status on 5 October 2026, two months after the paper appeared.
Lilly's 16 mg trial: the paper and the registry
Lilly's own phase 2 trial was published in The Lancet Diabetes & Endocrinology on 21 August 2026 (Hsia SH, Bays HE et al.), with its primary endpoint at 32 weeks: −18.1% at 16 mg, −15.6% at 10 mg and −7.3% at 3–6 mg against −0.9% on placebo, by the efficacy estimand. It is also the only mazdutide study with results on ClinicalTrials.gov, and the registry adds the 48-week figures and the raw counts behind the safety story:
| Lilly NCT06124807, week 48 | Placebo | 3/6 mg | 10 mg | 16 mg |
|---|---|---|---|---|
| Mean weight change | −0.05% | −10.54% | −19.2% | −22.3% |
| Started / completed | 48 / 29 | 32 / 28 | 48 / 37 | 51 / 34 |
| Stopped for an adverse event | 1 | 0 | 2 | 5 |
| Nausea | 26% | 44% | 47% | 67% |
| Vomiting | 2% | 22% | 13% | 47% |
Registry results module, first posted 21 April 2026; percentages are our arithmetic from the posted counts in the safety population (47, 32, 47, 51).
The 16 mg arm lost more than a fifth of body weight in a year, and paid for it in tolerability: two in three reported nausea, nearly half vomited, and a third did not finish. The two records count stopping differently: the paper reports that 20% at 16 mg discontinued treatment because of adverse events, while the registry's participant flow lists 5 of 51 leaving the study for that reason — people can stop the drug and stay in follow-up. A separate Lilly phase 1 study took 24 people to 16 mg over 20 weeks and reported −20.0% and −21.0% in two escalation cohorts (Diabetes, Obesity and Metabolism, 2025). Lilly has also registered a phase 2 trial in alcohol use disorder (NCT06817356, 308 participants, completed March 2026, no results posted).
Side effects
The pattern is the class pattern, sharper at higher doses. In GLORY-2 at 9 mg: vomiting 53.1% against 1.3% on placebo, nausea 46.9% against 3.2%, diarrhoea 39.4% against 6.5%, mostly mild to moderate, with 2.9% stopping for an adverse event against none on placebo. In GLORY-1 at 4 and 6 mg, discontinuation for adverse events was 1.5% and 0.5% against 1.0% on placebo. How the GLP-1 class's gastrointestinal effects compare is on GLP-1 side effects.
What cannot be said yet
- No US or European approval and no announced filing. Lilly's registered trials are phase 1 and 2.
- No cardiovascular outcomes trial is registered by either company; a Chinese investigator study of coronary plaque (NCT07657676, 116 participants) has not started.
- No trial against tirzepatide. The head-to-heads are against semaglutide and dulaglutide, in Chinese adults with type 2 diabetes.
- Almost all efficacy data come from Chinese adults, with lower average body weight than Western trial populations (GLORY-2: 94.0 kg, BMI 34.3). Lilly's 179-person US trial is the largest controlled dataset outside China.
Mazdutide offered for sale online outside China is not an approved medicine; FDA's position on unapproved GLP-1 products is on can you buy a GLP-1 over the counter?. The other pipeline drugs are on retatrutide, CagriSema and survodutide; approved drugs are compared on the best GLP-1 for weight loss.
Sources
Ji L et al. (GLORY-1), N Engl J Med 2025;392:2215–2225, DOI 10.1056/NEJMoa2411528 · Gao L et al. (GLORY-2), JAMA 2026;336:377–388, DOI 10.1001/jama.2026.8142 · Zhu D et al. (DREAMS-1), Nature 2026;652:174–180 · Guo L et al. (DREAMS-2), Nature 2026;652:181–188 · Hsia SH, Bays HE et al., Lancet Diabetes Endocrinol 2026, DOI 10.1016/S2213-8587(26)00160-9 · Bhattachar SN et al., Diabetes Obes Metab 2025;27:6460–6469 · Innovent Biologics announcements of 27 June 2025, 19 September 2025, 27 October 2025 and 25 November 2025 (via BioSpace) · ClinicalTrials.gov API v2: all studies listing mazdutide, and the results module of NCT06124807 · openFDA Drugs@FDA query. All read 5 October 2026.