BGM0504
BGM0504, dual GLP-1 and GIP receptor agonist
Written by Aaron CuhaReviewed Sep 2026
Also known as: BGM0504 injection, BGM0504 tablets
A tirzepatide-derived dual agonist that beat semaglutide 1 mg on glycated haemoglobin and weight in a 64 person phase 2. Its phase 3 obesity result, widely quoted at 19.3 percent weight loss, exists only as a company press release and has not been published.
Overview
BGM0504 comes from BrightGene Bio-Medical in Suzhou and is explicitly derived from tirzepatide. The published design paper says so: molecular dynamics simulations identified a salt bridge between the parent peptide and both receptors, the acylation side chain was repositioned on that basis, and the resulting molecule showed two to three times the agonist activity of tirzepatide in vitro while keeping a long plasma half-life.
That is a real piece of protein engineering and the paper is published in a peer-reviewed journal. It is also, precisely, an in vitro and animal potency claim. Higher receptor activity in a cell assay is not a clinical result.
The human data published so far is a phase 1 dose escalation in 40 healthy volunteers and a phase 2 in 64 Chinese adults with type 2 diabetes. In the phase 2 the 15 mg dose lowered glycated haemoglobin 2.48 percentage points against 1.43 for semaglutide 1 mg, and showed greater weight reduction than semaglutide. With 12 people per arm, those are signals. The authors use exactly that word.
Where it gets awkward is the phase 3. In May 2026 the company announced by press release that its phase 3 obesity trial met its primary endpoint with 19.3 percent mean weight reduction, a 16.5 cm waist reduction, blood pressure and uric acid improvements and a 0.7 percent discontinuation rate in the high-dose group. As of this review none of that had been published in a peer-reviewed journal. It is a company statement about its own unpublished trial, and it is the number every secondary source quotes.
BrightGene also has a separate long-acting compound, BGM1812, documented on this site on its own page.
Mechanism of action
A dual agonist at the GLP-1 receptor and the glucose-dependent insulinotropic polypeptide receptor, the same receptor pair tirzepatide acts on, engineered from the tirzepatide scaffold. The published optimisation work used molecular dynamics simulation to identify a salt bridge between the parent peptide and both receptors at residue K20 that was not visible in the cryo-electron microscopy structures, then repositioned the fatty acid acylation side chain to preserve it. The reported result was a twofold to threefold increase in agonist activity against tirzepatide in vitro with the extended plasma half-life retained. GLP-1 receptor activation reduces appetite and improves glycaemic control; the contribution of GIP receptor agonism to weight loss remains contested in this class, with agonists and antagonists both in development.
Human evidence
A 40 volunteer phase 1 and a 64 participant phase 2 in type 2 diabetes, both published. A phase 3 obesity result was announced by press release in May 2026 and has not been published.
- Phase 2, 12 weeks, 12 participants per BGM0504 arm: glycated haemoglobin fell 2.48 percentage points on 15 mg against 1.43 on semaglutide 1 mg, with greater weight reduction at the top dose.
- Phase 1, healthy volunteers: weight fell 3.24 to 8.30 percent across the dose range over a short titration period, with dose-linear pharmacokinetics.
- The company announced in May 2026 that its phase 3 obesity trial met its primary endpoint with 19.3 percent mean weight loss and a 16.5 cm waist circumference reduction. No peer-reviewed publication of that trial was located for this review.
- Multiple phase 3 trials in type 2 diabetes and obesity are recruiting in China, one in Indonesia, and a tablet formulation is in phase 1.
What this does not tell you: The published human evidence totals 104 people across two trials, the longer of which ran 12 weeks. The comparison that makes the compound interesting, against semaglutide, was 12 participants per arm. The design paper's central claim, two to three times the potency of tirzepatide, is an in vitro measurement and has never been tested against tirzepatide in a human trial. And the phase 3 numbers in general circulation are a company press release, not a published trial.
Reading the research record
This entry exists partly to be explicit about a distinction that gets blurred constantly in this field: the difference between a published trial and an announced one.
The 19.3 percent weight loss figure attached to BGM0504 comes from a press release issued by the company that makes it, about a trial it ran, which has not been through peer review and whose protocol details, dropout handling, estimand and adverse event table are not publicly available. It may well hold up. Press-released phase 3 results usually do. But until it is published, nobody outside the company can check how weight loss was calculated, who was excluded from the analysis, or what the completer rate was.
The published record is smaller and more modest: a 40 person phase 1 and a 64 person phase 2 whose authors describe their own findings as a signal.
The in vitro potency claim deserves its own caution. Two to three times the receptor activity of tirzepatide in a cell assay does not predict two to three times the clinical effect, or any clinical advantage at all. Potency determines the dose, not the ceiling. Whether this molecule does more than tirzepatide in people is an untested question, because no trial has compared them.
One practical warning applies to every compound on this page. Chinese incretin drugs carry several names at once: a company development code, a Chinese trade name, an international nonproprietary name assigned later, and occasionally a partner's code after a licensing deal. Those names do not reliably appear together in any one source, so a search on the code can return only preclinical work while the clinical trials are indexed under the nonproprietary name, or the reverse. If a compound looks like it has thin evidence, check the other names before concluding that.
The evidence, charted
Fig. 1 · evidence composition
2of 3 citations (67%) are in people
Counted from the citation list on this page. The Human count is the same number shown in the badge at the top. Both understate any literature larger than the sources we cite.
Fig. 2 · evidence over time
Evidence spans 3 distinct years, 2024 to 2026, counted from the citation list on this page.
Fig. 3 · legal status at a glance
US
Not approved
UK
Not approved
AU
Not approved
CA
Not approved
Not approved in any of the four jurisdictions shown. A jurisdiction's classification is a regulatory fact, not a verdict on the science; see Legal status below for the full text and any notes.
Fig. 4 · dose response
No human dose response curve exists
We draw this figure where the data supports it. For this compound in humans it does not, so the panel stays empty rather than borrowing an animal curve and implying it transfers.
Key studies & citations
- Human2026
Evaluation of the Safety, Efficacy and Pharmacokinetics of BGM0504 in Chinese Adults With Type 2 Diabetes: A Multicentre, Randomised, Controlled and Double-Blind Phase II Trial
64 Chinese adults with type 2 diabetes randomised to BGM0504 at 5, 10 or 15 mg, placebo, or semaglutide 1 mg, all weekly for 12 weeks. Mean glycated haemoglobin change was minus 1.72, minus 1.94 and minus 2.48 percentage points against minus 1.43 for semaglutide and plus 0.28 for placebo. The 15 mg dose showed greater weight reduction than semaglutide. The authors describe the findings as a signal throughout, which is appropriate for 12 participants per arm over 12 weeks. Five authors are BrightGene employees who may hold stock.
Diabetes, Obesity and Metabolism - Human2025
The safety, tolerability, pharmacokinetics and pharmacodynamics of an optimized dual GLP-1/GIP receptor agonist (BGM0504) in healthy volunteers: A dose-escalation Phase I study
40 healthy Chinese volunteers, single 2.5 mg dose and weekly titration to 5, 10 and 15 mg. Maximum concentration and area under the curve were linearly proportional to dose across 2.5 to 15 mg, supporting weekly dosing. Body weight change from baseline was minus 3.24, minus 6.26, minus 7.09 and minus 8.30 percent at 2.5, 5, 10 and 15 mg. Gastrointestinal events, vomiting, nausea, decreased appetite, diarrhoea and abdominal distension, were the most frequent adverse events. These weight figures are from healthy volunteers over a short dosing period, not from people with obesity.
Diabetes, Obesity and Metabolism - In vitro2024
Molecular dynamics-guided optimization of BGM0504 enhances dual-target agonism for combating diabetes and obesity
The design paper. Molecular dynamics simulation identified a salt bridge between the tirzepatide parent peptide and both receptors at K20 that cryo-electron microscopy structures did not show, and the acylation side chain was repositioned accordingly. The optimised peptide showed a twofold to threefold increase in agonist activity over tirzepatide in vitro while retaining extended plasma half-life, and outperformed tirzepatide in animal studies. This is a potency and design result in cells and animals; it makes no clinical claim.
Scientific Reports
Frequently asked questions
Is BGM0504 better than tirzepatide?
Unknown, and no trial has compared them. The claim of two to three times greater agonist activity comes from cell assays in the design paper. Greater potency in vitro sets the dose, not the maximum achievable effect, and it does not establish any clinical advantage.
Where does the 19.3 percent weight loss number come from?
A company press release issued in May 2026 announcing that the phase 3 obesity trial met its primary endpoint. It has not been published in a peer-reviewed journal, so the trial's analysis population, dropout handling and full safety table are not publicly checkable.
What does the published human evidence actually show?
A phase 1 in 40 healthy volunteers with dose-linear pharmacokinetics and 3 to 8 percent weight change over a short titration, and a phase 2 in 64 adults with type 2 diabetes where the 15 mg dose lowered glycated haemoglobin 2.48 percentage points against 1.43 for semaglutide 1 mg over 12 weeks. That is 104 people in total.
Can I buy it?
It is not approved in any country. Phase 3 trials are recruiting in China and Indonesia. Material sold online under this code is not the trial product and has no verified identity or purity.
Is it related to BGM1812?
Same company, different molecule and different programme. BGM1812 is documented separately on this site.