Mazdutide Dosage: The Phase 3 Doses, the 9 mg Trade-Off & Reconstitution Math

Mazdutide at a glance

What it is

Mazdutide (development codes IBI362 and LY3305677), a once-weekly glucagon-receptor and GLP-1-receptor dual agonist. Indexed as MeSH supplementary concept C000719829, which lists mazdutide, IBI362 and LY3305677 as the same substance. No molecular formula or mass is published on this page: PubChem's REST service returned a server-busy fault on every attempt while this page was written, and an unverified formula is worse than none

Researched for

Long-term weight management and type 2 diabetes — both approved indications in China — plus ongoing clinical evaluation in metabolic dysfunction-associated fatty liver disease, obstructive sleep apnoea and alcohol use disorder

Published trial doses

4 mg and 6 mg once weekly in GLORY-1 and in both DREAMS phase 3 trials; 9 mg once weekly in GLORY-2; 3 mg, 4.5 mg and 6 mg in the phase 2 programme; a 32-person phase 1 escalated to 16 mg. No trial started a participant at its maintenance dose

Route reported

Subcutaneous injection, once weekly, in every published trial in the programme

Reported frequency

Once weekly. Not a daily compound

Reported cycle

No cycling in any trial. GLORY-1 dosed continuously for 48 weeks and GLORY-2 for 60 weeks

Half-life

Not established in the indexed literature. A PubMed search pairing mazdutide — or either development code, which return the same record set — with pharmacokinetics or half-life terms returns no records at all, while the same database holds 48 records on the compound overall. What the record does establish is the interval: every phase 2 and phase 3 trial dosed once weekly

Regulatory status

Approved in China: June 2025 for long-term weight management in adults with a BMI of at least 28, or at least 24 with a weight-related comorbidity, and September 2025 for glycaemic control in type 2 diabetes. Marketed as Xinermei. Not approved in the United States or the European Union. Sold and referenced here for research use only

Reported ranges from research/community — examples, not recommendations.

What it is / mechanism

Mazdutide is a synthetic peptide that activates two receptors at once: the glucagon receptor (GCGR) and the glucagon-like peptide-1 receptor (GLP-1R). It is built on the oxyntomodulin scaffold — oxyntomodulin is the naturally occurring gut hormone that happens to hit both of those receptors — which is a different design lineage from tirzepatide, whose second receptor is GIP rather than glucagon. The two development codes are the single most common source of confusion about this compound, and they are not two molecules. IBI362 is Innovent Biologics' code and LY3305677 is Eli Lilly's; the two companies co-developed it, and the MeSH supplementary concept record C000719829 lists mazdutide, IBI362 and LY3305677 as names for one substance. A vial labelled IBI362 and a vial labelled LY3305677 are labelled for the same thing, which is worth knowing before paying a premium for one designation. The glucagon arm is the part that inverts most people's intuition. Glucagon is known as the hormone that raises blood glucose, so adding glucagon-receptor agonism to a diabetes drug reads like a mistake. The dual-agonist rationale is that glucagon-receptor signalling contributes increased energy expenditure and direct effects on hepatic fat handling, while the GLP-1 arm supplies appetite suppression, slowed gastric emptying and glucose-dependent insulin secretion — and offsets the glycaemic effect of the glucagon arm. The DREAMS-1 result is the cleanest evidence that the offset works in practice: mazdutide monotherapy lowered HbA1c by 1.57 to 2.15 percentage points against 0.14 on placebo, which is not what a net glucagon agonist would do. That hepatic mechanism is also why the programme extends past obesity and diabetes into metabolic dysfunction-associated fatty liver disease, obstructive sleep apnoea and alcohol use disorder — indications a pure appetite-suppression mechanism would not obviously reach.

Researched effects

Mazdutide is one of the few compounds in this library where the answer to "what does it do" comes from phase 3 trials that a regulator has already reviewed, so the numbers are worth quoting rather than summarising. GLORY-1 (NCT05607680, published in the New England Journal of Medicine in 2025) randomised 610 Chinese adults with a BMI of at least 28, or 24 to under 28 with a weight-related coexisting condition, to 4 mg mazdutide, 6 mg mazdutide or placebo for 48 weeks. Mean baseline weight was 87.2 kg and mean BMI 31.1. At week 32 the mean change in body weight was -10.09% on 4 mg and -12.55% on 6 mg, against +0.45% on placebo, and 73.9%, 82.0% and 10.5% of participants respectively had lost at least 5%. By week 48 the figures were -11.00% and -14.01% against +0.30%, with at least 15% lost by 35.7%, 49.5% and 2.0%. GLORY-2 (NCT06164873, published in JAMA in 2026) is the trial that tested the higher dose. It randomised Chinese adults with a BMI of at least 30 in a 2:1 ratio to 9 mg mazdutide or placebo for 60 weeks across 27 hospitals; 461 participants were analysed, mean age 33.9, mean BMI 34.3, 16.1% with type 2 diabetes. At week 60 the mean change in body weight was -16.65% against -1.50% on placebo, a between-group difference of -15.15 percentage points, and 84.3% versus 33.1% had lost at least 5%. Two things in that second result deserve more attention than the headline. The first is that 33.1% of the placebo group lost at least 5% of body weight — both trials ran alongside a reduced-calorie diet and increased physical activity, and in GLORY-2 that background alone moved a third of the placebo arm past the 5% threshold. The second is the cost of the 9 mg dose, which is in the safety section below and is the part of this compound's record most often left out. In type 2 diabetes the two DREAMS phase 3 trials reported separately. Against placebo (320 participants, 24 weeks) mazdutide lowered HbA1c by 1.57 percentage points at 4 mg and 2.15 at 6 mg, against 0.14 on placebo, with weight changes of -5.61% and -7.81% against -1.26%. Against dulaglutide 1.5 mg (731 participants, 28 weeks) both doses were superior on HbA1c by 0.24 and 0.30 percentage points, and on body weight by 3.78 and 5.76 percentage points.

Dosage — reported ranges (overview)

The doses with phase 3 evidence behind them are 4 mg, 6 mg and 9 mg once weekly. That is an unusually firm statement for this library, because these are not figures reconstructed from forum convention — they are the doses a regulator reviewed. The first thing the trial record establishes is that nobody starts at those numbers. The phase 2 programme worked upward through 3 mg, 4.5 mg and 6 mg; the high-dose phase 1 used two separate escalation regimens spread over 20 weeks to reach 16 mg. Escalation is not a detail of the protocols, it is how every one of them was run, and a page that quotes 9 mg without saying it was arrived at rather than started at has quoted half a fact. The second thing is the shape of the dose-response curve, and it is not linear in the direction buyers usually assume. In the phase 2 obesity trial, going from 3 mg to 4.5 mg bought 3.7 percentage points of weight change (-6.7% to -10.4%); going from 4.5 mg to 6 mg bought 0.9 (-10.4% to -11.3%). The curve was already flattening inside phase 2. The third is the 9 mg question, and it needs a caveat before the numbers. GLORY-1's 6 mg arm reached -14.01% at week 48; GLORY-2's 9 mg arm reached -16.65% at week 60. Those are different trials, different durations, and different populations — GLORY-2's participants had a mean BMI of 34.3 against GLORY-1's 31.1 — so the difference between them is not a clean measurement of what 9 mg adds over 6 mg, and no head-to-head trial of the two doses has been published. What can be read cleanly is what happened inside GLORY-2 on its own: 9 mg produced a 15.15-point separation from placebo, and it did so in a trial where 53.1% of participants vomited against 1.3% on placebo. None of this is a protocol, and none of it is a recommendation. It is what the published trials administered, in what order, with what result.

A printable protocol sheet with a reconstitution reference and an injection log comes with All-Access Lifetime.

Reconstitution — bac-water math

Research-grade mazdutide is supplied lyophilised and dosed in whole milligrams, which keeps the concentration arithmetic simple and makes the syringe arithmetic the part that actually constrains you. Concentration in mg/mL is the vial's milligrams divided by the millilitres of bacteriostatic water added. On a U-100 insulin syringe 100 units is 1 mL, so units to draw = dose in mg ÷ concentration in mg/mL × 100. Worked example on a 10 mg vial: add 1 mL and the concentration is 10 mg/mL, so a 4 mg example is 4 ÷ 10 × 100 = 40 units and a 6 mg example is 60 units. The 9 mg dose is where dilution choices start to bite. At 10 mg/mL it is 90 units — still one syringe, but only just. Add 1.5 mL to the same vial instead and the concentration falls to about 6.67 mg/mL, at which point 9 mg is 135 units and no longer fits in a single U-100 draw. The rule that follows is the opposite of the intuition that more diluent is gentler: for a compound dosed in whole milligrams, over-dilution is what pushes a dose past the syringe. This is concentration math, not a dose. The figures below are arithmetic worked on the doses the trials published.

This is concentration math, not a dose recommendation.

Injection / administration basics

Every mazdutide trial in the registry used subcutaneous injection once weekly, so there is no ambiguity about route or schedule here — the published protocols and the market convention agree, because the convention was copied from the protocols. What the trials describe in practical terms is a weekly subcutaneous administration inside a titrating regimen, with escalation measured in weeks to months rather than days. GLORY-1 ran 48 weeks and GLORY-2 ran 60, both continuously, both alongside a reduced-calorie diet and increased physical activity that the protocols treated as part of the intervention rather than as background noise — in GLORY-2 that component alone moved a third of the placebo arm past 5% weight loss. One practical consequence of a once-weekly compound is that a reconstituted vial has to stay stable for weeks rather than days, which makes the diluent choice and the storage discipline matter more than they would for a daily peptide. That is the reason bacteriostatic water rather than sterile water is the conventional choice for multi-dose vials: the benzyl alcohol is a preservative for repeated entry, which a single-use sterile-water preparation does not provide. Route and schedule are reported here as published trial parameters. Nothing on this page is an administration instruction.

Half-life & frequency rationale

No half-life for mazdutide is established in the indexed literature, and this is worth stating precisely rather than filling with a plausible number. A PubMed search pairing mazdutide with pharmacokinetics or with half-life terms returns no records. The same search run against both development codes returns the same answer, because IBI362 and LY3305677 retrieve the same record set as the generic name — 49 records, none of which is a pharmacokinetic study. That is not a search artefact: the identical query structure pairing mazdutide with obesity returns 39 records, so the database is answering the question rather than failing to parse it. What the record does establish is the interval, which is the practically relevant quantity anyway. Every phase 1b, phase 2 and phase 3 trial in the programme dosed once weekly, and the regulatory approval is for a once-weekly product. A dosing interval that survived four phase 3 trials is a stronger statement about duration of action than a half-life figure copied from a vendor page would be. If you see a specific half-life quoted for this compound, the useful question is which study it came from. We could not source one.

Stacking — overview

No published trial has combined mazdutide with another peptide, and there is a specific reason to distrust the pairings the market suggests. Mazdutide already contains a GLP-1 receptor agonist arm. Pairing it with semaglutide, tirzepatide, retatrutide or cagrilintide is not additive in the way stacking usually implies — it is activating the same receptor twice, on a compound whose gastrointestinal adverse-event rate reached a majority of participants at 9 mg by itself. The interesting relationships in this class are comparisons rather than combinations, which is how they are framed below. One genuine comparison does exist in the trial record rather than in speculation: DREAMS-2 ran mazdutide head-to-head against dulaglutide, a single-receptor GLP-1 agonist, and mazdutide was superior on both HbA1c and body weight. That is a published comparison between mechanisms, and it is the only head-to-head this compound has.

For how combinations are grouped by research context, the named blends, and why a pre-mixed blend vial cannot be calculated from its total milligrams, see the peptide stacks guide.

Storage & handling

• Lyophilized (powder): store sealed and refrigerated at 2-8°C, protected from light and moisture; long-term storage is commonly at -20°C or below. Do not use past intended research shelf life.
• Reconstituted (liquid): refrigerate at 2-8°C, protect from light, and avoid freeze-thaw cycles and vigorous shaking.
• A once-weekly compound means a reconstituted vial is entered repeatedly over weeks, which is the case bacteriostatic water exists for — its benzyl alcohol is a preservative for multi-dose entry that sterile water does not provide.
• Label the vial with the reconstitution date and the resulting concentration. The concentration is the number the unit arithmetic depends on, and it is the one nobody remembers a fortnight later.

Guide FAQ

Quick answers about guide scope, access, and educational use context.