I started researching peptides because I was turning fifty
That number does something to how you think. Reversing damage is hard. Slowing it down is achievable — but only while there is still something worth preserving. I could see the window narrowing, and I wanted to spend what was left of it on the right side of the problem.
So I did what I do with everything: went all the way in. Read the literature. Read the patents. Talked to people who formulate drugs for a living. Bought product from most of the well-known names to see what I was actually getting. What I found made me build this lab.
It exists. Almost nobody is using it.
Here is what genuinely surprised me: the information is out there. Published clinical trials with real dosing arms and real endpoints. Patents — often the single richest source of formulation detail available, because to claim something you have to describe it in enough detail that someone else could reproduce it. Pharmacokinetic studies, dose-escalation data, receptor binding constants, half-lives, bioavailability by route. Regulatory submissions and approved labels for the peptides that made it to market.
Almost none of it shows up in how these compounds are actually sold and used.
What circulates instead is a number somebody posted once. It gets repeated on a forum. It gets quoted in a video. It shows up on a vendor's dosing chart. Six months later it is “the standard protocol,” and if you ask where it came from, the answer is that everyone knows it. The number has no origin. It was never derived from anything. It accumulated authority through repetition.
Same with mechanism. You will see a compound described as working through a pathway it does not meaningfully touch, at a dose that would not engage that pathway if it did. Not because anyone is lying — because the claim was copied from someone who copied it from someone else, and nobody in the chain went back to the source.
I got in the habit of checking. Pick any widely repeated protocol, trace it backward, and you can usually find the exact point where a guess became a fact. It is rarely more than two or three steps. That is the first thing this lab was built to reject — every dose and every ratio we use traces back to something. If I cannot trace it, I do not build on it.
You have no idea who made what is in your vial
The second thing I learned took longer, and it bothered me more. The overwhelming majority of companies selling peptides are not making them. They buy finished or near-finished product from a small number of contract manufacturers, apply their own label, and resell. Follow the supply chain of a dozen well-known brands and it collapses into a handful of actual sources — sometimes fewer.
Which means when you buy from a name you trust, you are usually not buying that company's manufacturing. You are buying their marketing, wrapped around somebody else's vial. And that raises questions nobody can answer for you:
- What facility was this made in — a classified cleanroom, or a room someone calls a cleanroom?
- How was it filled? Under inert gas, or in open air?
- What were the vials and stoppers treated with before filling — anything at all?
- What water was used? Water for injection with a certificate, or something less?
- What grade were the excipients? USP, or whatever was cheapest that dissolved?
- Is there a batch record? Could anyone reconstruct how this specific vial was made?
- Who touched it, in what garment, in what air?
For most product on the market those questions have no answer available at any price. Not a bad answer — no answer. The reseller does not know either.
And it gets worse than “unknown”
The FDA states it plainly: counterfeit drugs claim to be authentic but could contain the wrong ingredients, too little, too much, or no active ingredient at all. Drug-safety organisations tracking this describe online sites selling substances labelled as semaglutide, tirzepatide and retatrutide, often sourced from foreign factories not registered with the FDA. These chemicals, in their words, have no pedigree.
The enforcement volume tells you the scale. Dozens of warning letters to online sellers, and adverse-event tracking recording more than 455 reports tied to compounded semaglutide and more than 320 tied to compounded tirzepatide as of early 2025. Published analysis has found a substantial share of compounded tirzepatide samples testing below claimed purity — which means people are running protocols, adjusting doses and drawing conclusions about how a compound works for them, based on a vial whose actual contents nobody ever measured.
Mannitol is a bulking agent. Its entire job is to build the white disc of cake in the bottom of the vial — it is what makes a lyophilised product look finished. Vendors are currently shipping vials that are mannitol and nothing else. No peptide. Not underdosed — absent.
Think about what that means mechanically. The cake looks perfect. It has to look perfect, because the filler that makes it look perfect is the only thing in there. It reconstitutes clean. It draws into a syringe like anything else. There is no visual inspection, no smell, no texture, no dissolution behaviour that distinguishes a vial of drug from a vial of sugar. Fake and real look identical, and the fake one arguably looks better.
It also explains something people misread constantly. Someone runs a compound for six weeks, feels nothing, and concludes the compound does not work. Posts about it. Others repeat it. Now there is a forum consensus that a molecule with published trial data is useless — built on a batch of vials that never contained it.
What independent testing keeps finding
The most-cited number in this market comes from Janoshik Analytical, the Czech lab the research-peptide community actually uses: of the peptides submitted to them in 2024, 43% failed to meet their own label purity claims. The breakdown is the useful part — lower-tier vendors came back at actual purities of 71–91% against claimed 99%+. Not counterfeit, not empty. Just quietly short, by amounts nobody would notice from the outside.
Two honest qualifications on that number, because this page would be a hypocrite without them. That figure includes private and withdrawn submissions, so read it as an order-of-magnitude signal rather than a precise statistic. And Janoshik is not ISO 17025 accredited — its public database shows only the batches vendors chose to submit, which if anything selects for the ones expected to pass. Take it as what it is: the best measurement available in a market that has no other, pointing at a failure rate somewhere in the tens of percent.
What is actually in the vial is usually mannitol
Then I started looking at the formulations themselves, and this is where it stopped being a supply chain problem and became a pharmaceutical one. Open the typical research peptide vial and the ingredient list is the peptide and mannitol. Frequently that is all of it.
Mannitol builds the cake. It is not a stabiliser in any meaningful sense — it crystallises during freeze-drying and phase-separates from the peptide, so the molecule ends up pushed out of the matrix rather than protected inside it. Comparative studies consistently find amorphous sugar stabilisers outperform it. It exists in that vial because it produces a good-looking cake cheaply.
More importantly: mannitol has no buffering capacity at all. So when you add bacteriostatic water, the pH of that solution goes wherever the peptide's own residual chemistry drags it. Nothing in the vial has any say in the matter.
That matters enormously, because peptides are made by solid-phase synthesis and purified by reverse-phase HPLC, and both steps use trifluoroacetic acid. The peptide comes off the line as a TFA salt. For pharmaceutical use that counterion is routinely exchanged for acetate, because acetate is non-toxic and trifluoroacetate is not — published work reports it as cytotoxic at low concentrations, and as interfering with cell proliferation, receptor binding and enzyme activity. The exchange costs money and loses yield, so it is the first step skipped. Freeze-drying will not save you: it removes free TFA but not the counterion bound to the peptide itself.
Put an unexchanged TFA-salt peptide in a vial containing nothing but mannitol, add water, and the result is an acidic, unbuffered, unprotected solution. That lines up closely with what people commonly report:
- Stinging, redness and knots at the injection site. Acidic pH and residual trifluoroacetate are both documented irritants.
- Faster loss of potency. Hydrolysis and deamidation rates are strongly pH-dependent, so an off-target pH places a peptide outside its published stability window from the moment it is reconstituted.
- Cloudiness, stringing, gelling. Solubility falls sharply near a peptide's isoelectric point, and an unbuffered solution can land there.
Pharmaceutical formulation exists because the formula determines whether the active ingredient works. That is not a nicety layered on top of the drug. It is part of the drug.
What this is, and what it is not
What this is: a formulation and research operation. We develop peptide formulations, we do the science behind them, and we build the processes those formulations need in order to survive contact with the real world.
What it is not: a manufacturer. We are not trying to become a company that makes everything. Other companies may end up producing formulas we develop, and that is fine — the formula is the work. Selling what we formulate is a by-product of doing the research, not the reason for doing it. The goal is to create the future of peptide science; everything else exists in service of that.
Which meant building a facility capable of doing formulation work properly, because you cannot develop and validate a formulation in an environment that introduces more variables than it controls. I built it to do things the pharmaceutical way, and where the pharmaceutical way costs millions, I engineered around the cost instead of around the requirement.
None of that is exotic. It is what the good drugs get. It is simply absent from this market, because it is slow, it costs money, and nothing forces anyone to do it. And it keeps getting better — the room, the procedures and the formulas are under continuous improvement, because the standard I am building toward is the one used for drugs that cost a thousand dollars a vial.
Formulating for how the body actually works
Once the lab could produce a clean, stable, properly formulated vial, the interesting question opened up: what should be in it?
Almost everything sold in this space is single-compound. One peptide, one vial, one purpose. But that is not how any of these pathways operate. Biology runs on cascades and feedback loops, and compounds that touch different points in the same cascade do not simply add — they can multiply, or cancel, depending entirely on ratio and timing.
So development here has three inputs. The literature — human trial data where it exists, mechanistic work where it does not, receptor binding constants, pharmacokinetics, and the dose-response curves that show where a compound stops doing more and starts doing something else. Patents get read closely, because they frequently carry formulation and ratio detail found nowhere in the published literature.
People who do this professionally. I have consulted formulators, analytical chemists and clinicians who work on real drugs. Every one of those conversations changed something. The single most valuable thing about talking to someone who formulates for a living is how quickly they tell you what will not work, and why.
Trial and error, run honestly. Small pilot batches. Reconstitution testing. Watching how a formulation behaves during filtration, because filter loading and retention reveal solubility and aggregation problems before they ever reach a vial. Iterating on cake structure until dissolution is effectively instant — the formulation literature identifies agitation and air-liquid interface exposure as stress factors, so a cake that needs prolonged swirling is introducing something worth avoiding.
When we find a way to make a formula better, the formula changes. Nothing here is frozen because it is already printed on a label. And every formula is developed for its specific molecule — there is no house vehicle poured over everything in the catalogue. A chelator removes the trace metals research identifies as oxidation catalysts; put that same chelator in a copper-complexed peptide and it strips the metal that constitutes the active molecule itself. Same excipient, opposite outcome. Deciding what belongs in a given vial is the actual work.
I built this for myself first
That is not a slogan — it is a design constraint that shows up in every decision, because I am using the product and I would rather spend more and know exactly what I am putting in my body. What that produced:
- Every dose traceable to a source. Trials, patents, pharmacokinetic data. Nothing carried forward because it is what people say.
- A facility whose numbers are published. Air changes, filter classes, depyrogenation cycles, humidity and temperature targets. Check them.
- Pharmaceutical formulation, per peptide. Buffered, pH-adjusted, protected against oxidation and trace metals, stabilised for both the dry and the reconstituted state.
- A container engineered for the whole life of the vial — not just the moment it was sealed.
- Combinations designed to reduce injection burden wherever the chemistry genuinely allows it.
You do not have to trust any of that. That is rather the point. The process is documented, the parameters are stated, the reasoning is published, and independent testing of randomly selected vials is the strongest verification available precisely because it is the one I cannot influence. The market I walked into asked for trust and offered nothing to check. I would rather hand you something to check.
And the products are the smallest part of it. What I actually want to build is the science — formulations developed properly, processes that hold up, and a body of work that raises the floor for everyone in this field. If some of that reaches you as a vial, good. If it reaches you as something you can read, question, and use to evaluate anyone else's product, that is the point.
Two people is an anecdote. Twenty is a pattern — and the differences between them will teach you more than any single protocol can.
Read the protocol accounts