peppersite
Claim auditSemax2026-08-28 · 1,133 WORDS

Semax Is Approved in Russia. That's a Fact About a Regulator, Not a Result.

I went looking for the human trial I could hand you and came back with mice, a negative result, and one honest reason the Western data is thin.

MT
11 claims gradedAnimal

"Semax is approved in Russia."

That sentence arrives in my inbox in almost those exact words, usually bolted to a question about whether a 10mg vial is worth the money. I understand the logic. If a national regulator signed off, then somebody, somewhere, read a dossier. That feels like a shortcut past the tedious business of reading trials yourself.

We sell Semax 10mg. Purity is stated at ≥99%, the COA is attached, and our certificates confirm identity by LC-MS with the mass-confirmation spectra printed on them. Now let me undercut my own paperwork before anyone else does: that certificate tells you what is in the vial. It says nothing about whether the Semax literature supports what you want Semax to do. Purity is a manufacturing claim. It is not an evidence claim, and I have watched buyers read it as one.

So I went looking for the human trial I could put in front of you. I want to be precise about what I came back with.

The rung problem

Here is the lens I use for any compound whose research lives in a language I do not read. Call it the Imported Evidence Ladder, four rungs.

Bottom rung, community extrapolation: a claim on a forum that has borrowed credibility from somewhere higher up without ever standing there. Next rung, preclinical mechanism: rodent or test-tube work, which tells you how something might work and not whether it does. Above that, foreign clinical literature: actual humans, variable blinding and reporting, frequently never replicated in English. Top rung, a registered indication with a foreign regulator: a real dossier, foreign standards, and a population that is almost always narrower than the internet remembers.

A registration is a fact about a regulatory system. It certifies that an agency accepted a submission under its own rules for its own stated indication. It does not certify that the compound does what a healthy 34-year-old wants it to do at their desk on a Tuesday.

And I am not going to pretend I have read the dossier. I have not. It is not in English, I cannot cite it to you with a population and an endpoint, and quoting somebody else's summary of it as though it were data would be exactly the move this piece exists to criticize. So that top rung stays labeled and empty in my hands. That is not me dismissing the Russian clinical work. It is me refusing to spend credibility I did not earn.

What I can actually show you

Semax is a synthetic heptapeptide, seven amino acids, modeled on a fragment of ACTH called ACTH(4-10). ACTH is adrenocorticotropic hormone, the pituitary signal that tells your adrenal glands to make cortisol. Being built from a slice of a hormone tells you about the molecule's parentage. It does not tell you what the molecule does, and that gap is where most of the marketing lives.

The animal work is real and it is more interesting than the forum version.

In transgenic APPswe/PS1dE9/Blg mice, animals engineered to carry human genes that make them pile up amyloid plaques the way Alzheimer's brains do, Semax and a derivative improved performance on cognitive behavioral tests: open field, novel object recognition, and the Barnes maze, which is a table full of holes where the mouse has to remember which hole leads home. Histology backed it up. The treated mice had fewer amyloid inclusions in the cortex and hippocampus.

That is a genuinely good preclinical result. It is a disease-model result in engineered mice, which is the second rung, not the third.

Now the part vendor copy skips. In rodent models of Parkinson's-like disease, a qualitative review found that Semax did not significantly change striatal dopamine levels or improve locomotor activity, with one study reporting improved motor performance at a higher dose. The same review states flatly that no clinical trials have tested Semax's efficacy in human Parkinson's disease patients, and that more research is needed before it could be considered viable.

Stranger still: in one reviewed rodent study, Semax given alone produced no change in tissue or extracellular dopamine. Give it twenty minutes before d-amphetamine and it markedly amplified the amphetamine-driven dopamine release and locomotor activity. Alone, nothing. As an amplifier of a stimulant, plenty. If your mental model of Semax is "dopamine peptide," that finding should complicate it rather than confirm it.

The claim that maps to nothing

The healthy-user focus claim usually routes through BDNF, a growth factor that supports neurons. I do not have a BDNF study I can cite here, so I am not going to describe one, and I am certainly not going to walk a rodent gene-expression finding across the species barrier into your workday. That claim, as it circulates, sits on the bottom rung wearing the clinical literature's jacket.

Safety is the same shape. I have no human safety dataset with a duration and a population to report, so I will not be calling this compound well tolerated. Absence of reported harm in work I cannot cite is not a safety record.

Selank, and why animal data is slipperier than it looks

Semax's usual companion in these conversations is Selank, and Selank's rodent literature is a beautiful demonstration of how conditional this stuff is. Subchronic intranasal Selank in mice cut the number of NMDA receptor glycine binding sites in the cortex by 18% in BALB/c mice and 53% in C57Bl/6 mice. In the hippocampus, BALB/c showed a 15% increase and C57Bl/6 showed nothing. Switch to intraperitoneal injection and the pattern rearranges again: cortex down 24% and hippocampus down 11% in BALB/c, cortex down 15% and hippocampus down 45% in C57Bl/6.

Same peptide. Two strains of mouse. Effects that differ by a factor of three and sometimes point in opposite directions. Apparently mice have subcultures.

Elsewhere, rats given experimental thermal burns showed elevated cell-mediated immune parameters, and Selank at 100 μg/kg/day intraperitoneally for 14 days after burn modeling shifted white cell counts and neutrophil activity back toward control values. The authors read that as Selank acting on systemic immune disturbance caused by a local stressor. Interesting. Also rats with burns, which is a long way from anyone reading this.

What we owe you

Our Semax page says the compound is "studied in nootropic and neurotrophic research models," states purity, and attaches the COA. I stand behind that wording. But the argument I am making here, that you deserve to know when a compound's evidence base is foreign-registered, non-English, or preclinical only, commits us to labeling evidence tiers across the whole catalog and not just where it flatters us. We do not do that catalog-wide yet. I would rather write that sentence than imply otherwise.

View the Semax 10mg listing, purity ≥99%, COA attached, sold for laboratory research use only.

The takeaway

Before you accept any Semax claim, ask which population it came from. If the answer is a rodent disease model or a trial you cannot read, and the claim is about your focus at work, that claim has climbed a rung it did not earn. The most useful thing a foreign registration tells you is that somebody built a dossier for a narrow purpose, in another system, for someone who was already sick.

Evidence behind this piece
CLAIMS BY EVIDENCE LEVEL Human trials Human trials: not reported NOT REPORTED Animal Animal: 11 claims 11

Claims this piece draws on, by the strength of the research behind each one. Ordered strongest to weakest. Community discussion is shown separately because it indicates interest, not evidence.

Reconstitution calculator
Mass and diluent in, barrel reading out, with the rounding error named.

Question everything. Including me.

This is the discussion for the audit above. If the piece missed a study, over-read one, or graded a claim higher than the paper supports, say so here — a reply that cites its source gets read first.

One audit a month. No product emails.
I re-read a study everybody is citing and say what it actually supports, including when the answer costs me a sale.