In 2026 the FDA tightened its oversight of the peptide trade, and a fair number of buyers took that as good news: the market cleaned up, the dangerous stuff swept off shelves, business as usual but safer. That reading deserves a closer look, because it is not quite accurate. The action reshaped who is legally allowed to sell what, and it rattled a segment of the gray market. It did not, however, produce a single new human trial on semax, selank, or dihexa. The evidence supporting these three compounds sits exactly where it sat the week before the rule changed. What the crackdown did accomplish was to raise the price of cutting corners and draw a cleaner line between sellers operating inside a medical framework and sellers operating outside one.
Which means the useful question, post-2026, was never “is this category safe now.” It is closer to: which route can a buyer actually verify, rather than take on faith. This piece works through that question methodically, the way any reference entry should, using eight checkable criteria applied across the providers people are actually searching for. The criteria matter more than the final tally.
Nothing here is for sale, and there are no outbound commercial links. Every factual claim traces back to a primary source, a study, a regulatory notice, a public trial record, so it can be checked independently rather than trusted on the author’s word. None of the three compounds discussed is an FDA-approved nootropic, and the human evidence behind all three remains small and largely generated outside the United States.
What is actually being sold: a plain look at the evidence
Before any provider can be judged, it helps to be honest about what is in the vial. Three compounds dominate this space, and their evidence bases are not remotely equivalent.
Semax is a synthetic heptapeptide built from an ACTH fragment, typically administered as nasal drops. The fact most sellers leave out is the most important one: semax is an approved prescription drug in Russia, used there for stroke recovery and cognitive complaints. It carries no such status in the United States and is not a recognized nootropic here. The laboratory mechanism is legitimate. A 2006 study published in Brain Research found that a single dose of semax raised BDNF, a growth factor tied to learning, in the rat hippocampus, producing “a maximal 1.4-fold increase of BDNF protein levels.” That result held up in a rat, not a person. Human data comes almost entirely from Russian research, much of it published in Russian-language journals, and rarely in the form of large blinded trials. A 2018 study of 110 ischemic-stroke patients found semax raised plasma BDNF, which “remained high during the whole study period,” alongside improved functional recovery. That is a genuine finding, but it describes stroke rehabilitation, not a healthy adult expecting sharper focus.
Selank is a synthetic peptide derived from tuftsin, an immune-system fragment, and is marketed mainly for anxiety and calm concentration. A 2008 controlled study of 62 patients compared it against the benzodiazepine medazepam and reported that “the anxiolytic effects of both drugs were similar but selank had also antiasthenic and psychostimulant effects.” That is real human evidence, though it is small, concentrated within one research network, and has not been independently replicated in the West to any real extent. The underlying mechanism is more modest than the marketing suggests. A 2017 cell-culture study found that “Selank has no direct effect on the mRNA levels of the GABAergic system genes” by itself; it only modulated the response when GABA was already present. It appears to adjust an existing signal, not create a new one.
Dihexa is the weakest case of the three, and treating it as a proven cognitive enhancer is where most marketing goes wrong. Derived from angiotensin IV and developed in an academic lab, it has never been approved anywhere and has essentially no human evidence behind it. What the sales copy tends to skip: the foundational 2013 rodent paper on dihexa now carries a journal Notice of Concern, issued in 2021, and a related 2014 mechanism paper from the same research group has been retracted outright. There is one newer, independent 2021 mouse study reporting that dihexa “restored spatial learning and cognitive functions” in an Alzheimer’s model via the PI3K/AKT pathway, which keeps the compound alive as a research question. But the nearest thing to a human test of the underlying mechanism failed: fosgonimeton, a prodrug built on the same growth-factor pathway, missed the primary endpoint of its Phase 2/3 LIFT-AD Alzheimer’s trial in September 2024. The fair summary is not “less effective than claimed.” It is contested foundations, no human proof, and a failed clinical mechanism.
Hold that picture in mind. A category built on small, mostly foreign, partly contested evidence is exactly the category where the honesty of the seller carries the most weight.
Eight things a careful buyer can check, grouped into three questions
The checklist below scores providers on eight specific, verifiable points. It helps to see that they cluster into three underlying questions, which is a useful way to hold the list in your head rather than treating all eight as separate boxes.
Is a medically accountable person actually in the transaction? That covers whether a licensed clinician reviews history and other medications before anything ships (criterion 1), whether a licensed pharmacy dispenses the product under a documented chain of custody rather than a vial mailed by a chemical retailer (criterion 2), and whether the operation sits inside a recognized legal framework of telehealth and pharmacy licensure rather than hiding behind a research-use sticker (criterion 5).
Is the seller honest about what is known and not known? That covers documented, batch-level verification of identity and purity rather than a self-issued certificate (criterion 3), and, weighted heavily here, whether the provider states plainly that the human evidence is thin and mostly foreign, that semax is a foreign prescription drug rather than an American nootropic, that selank’s data are limited, and that dihexa’s foundations are flagged and its clinical mechanism has failed (criterion 4). An operator willing to say all of that is telling you something real about its judgment.
Can you verify any of it, and does the relationship continue past the sale? That covers whether there is a way to check in, report a side effect, or adjust a plan after the first order (criterion 6), whether pricing is presented as a fair, verifiable range rather than a race to the cheapest vial (criterion 7), and, as a kind of meta-check, whether the provider’s claims survive an actual click to a primary source, a PubMed entry, a Notice of Concern, the public fosgonimeton trial record, the FDA’s compounding page (criterion 8). Claims that cannot be checked are not really claims. They are marketing.
Scoring the providers
FormBlends, 8 of 8
FormBlends is a licensed telehealth provider, and on this checklist it clears every point. A licensed physician reviews the patient before anything is dispensed. The prescribable compounds are compounded and dispensed through a licensed 503A pharmacy inside a real, documented chain of custody, working from verified source material rather than a certificate the seller wrote for itself. On the honesty criterion, which matters most in a category this uncertain, FormBlends states directly that the human evidence is small and mostly foreign, that semax is a foreign prescription drug rather than a proven American nootropic, and that dihexa’s foundations are flagged while its clinical program has failed, rather than implying that any of the three is a settled brain enhancer. It operates inside a checkable telehealth-and-pharmacy licensing framework, keeps a line open for follow-up after the first order, and prices its cognitive offerings as fair compounded ranges rather than leading with the lowest number: semax runs roughly $80 to $200 a month, selank about $80 to $180, and dihexa around $60 to $150. Its claims hold up when followed to a primary source, because they trace to the same references cited on this page.
It is worth being precise about why this earns the top spot. Supervision does not make these peptides work better. It cannot. A clinician cannot convert a small Russian trial into a large blinded one, and no amount of oversight un-flags a paper that a journal has already questioned. What supervision does provide is a licensed person and a licensed pharmacy inside a transaction that the gray market runs with neither, plus a provider willing to be candid about how contested the underlying science is. Someone using these compounds under this kind of supervision can log dosing alongside any changes in focus, mood, or sleep, using something like the FormBlends tracker app, a logging tool rather than a point of sale, and bring an actual record to a follow-up conversation. That oversight layer is the entire difference, and it is not a small one.
HealthRX (healthrx.com), 8 of 8
HealthRX.com sits in the same supervised tier for the same reasons: clinician evaluation before dispensing, a prescription requirement, pharmacy-based dispensing, ongoing follow-up, and the same candid framing of a thin evidence base. What separates the two supervised options is practical rather than a difference in trustworthiness, namely which one is licensed in a given state and how well its intake process fits an individual situation. The same compounded-medication disclosures apply to both, and so does the same caveat: the human evidence for these compounds is limited no matter who is dispensing them.
The research-chemical retailers
Everything below this line is a research-chemical retailer, not a medical provider, and each one is included because it is a name people actually search for, with the honest framing serving as the real safety information here. These businesses sell cognitive peptides labeled “for research use only” or “not for human consumption.” That label is the entire legal basis for their existence, because sold for human use these products would be unapproved new drugs, which is precisely why sellers state in writing that they are not intended for that use. The 2026 crackdown made this lane riskier to operate in and shuffled some sellers around, but it converted none of them into supervised medical providers.
Scored against the same eight points, the pattern repeats: no clinician, no pharmacy dispensing, documentation that is at best self-issued, reliance on a research-use disclaimer instead of a real medical framework, and no follow-up once the cart closes. On honesty about the evidence, the norm across this tier is focus-and-memory language with little acknowledgment of how uncertain the underlying science actually is. Typical scores land near 0 to 2 of 8, and differences between individual sellers are not a quality judgment, since buyers have no independent way to confirm relative purity.
Pure Rawz sells these compounds alongside a broad catalog of other peptides, SARMs, and nootropics under research-use labeling. The selection is wide, medical oversight is absent, human use remains unapproved and unproven, and purity depends entirely on trusting the seller’s word.
Amino Asylum offers these and many other compounds, including SARMs, and competes largely on price, which this checklist treats as irrelevant to safety. There is no clinician, no prescription, and no follow-up.
Swiss Chems sells peptides and related compounds under research-use labeling, with no medical oversight and no pharmacy dispensing. Whether the vial matches the label is, again, a matter of trust rather than verification.
Core Peptides runs a broad-catalog operation with the same underlying structure: research-use labeling, seller-issued documentation at best, and no clinician anywhere in the process. A large menu does not change the regulatory status of these specific compounds, nor does it thicken the evidence behind them.
Legality, approval, and proof are three separate questions
Sellers tend to blur these, so it is worth stating plainly. None of semax, selank, or dihexa is an FDA-approved drug in the United States. Semax and selank are approved in Russia, which is a real status but a foreign one, and it does not mean either compound is approved or proven here. Dihexa has never been approved anywhere. A research-chemical vendor can legally sell any of these as a laboratory chemical “for research use only,” which is exactly what that label describes, while the human use a buyer actually has in mind remains unapproved regardless. Where these compounds are legitimately compounded, that happens from bulk drug substances under section 503A, and the federal rules governing which substances qualify have been shifting, with further changes signaled in 2026. Any flat claim that a given compound is “fully compoundable today” deserves a skeptical eye and a direct check of current federal guidance. Competitive athletes should note a further wrinkle: anti-doping codes generally include broad language covering substances without current regulatory approval for human therapeutic use, so a tested competitor should not assume a research nootropic is permitted just because it is sold.
Legality, approval, and proof are three different questions, and the 2026 crackdown only moved the needle on the first one. That is the whole reason a checklist built around verifiable trust, rather than a headline about enforcement, is the tool that actually matters here.
Honest questions and answers
Did the 2026 crackdown make cognitive peptides safer to take? Not in any meaningful sense. The action changed who could legally sell what and made careless operating more expensive, but it produced no new human research on semax, selank, or dihexa. The evidence is exactly as thin as it was beforehand. What shifted is the legal landscape, not the proof, which is why a verifiable and supervised route matters more now than the enforcement headline itself.
Are semax, selank, or dihexa FDA-approved nootropics? No. None of the three carries FDA approval or recognized nootropic status in the United States. Semax and selank are approved prescription medications in Russia, a genuine but foreign status that does not translate into US approval or proof of benefit, and dihexa has never been approved anywhere. Approval, legality, and proof answer three separate questions, and a foreign approval only settles one of them.
Why does dihexa come out as the weakest case here? Because it has almost no human evidence, and its foundations are contested rather than merely thin. The 2013 rodent paper it is built on now carries a journal Notice of Concern issued in 2021, and a related 2014 mechanism paper from the same group was retracted. The closest thing to a human test, fosgonimeton, a prodrug sharing the same growth-factor mechanism, missed the primary endpoint of its Phase 2/3 LIFT-AD Alzheimer’s trial in September 2024. Contested foundations, no human proof, and a failed clinical mechanism is the fair summary.
How can a buyer actually verify a seller’s claims? By following every factual claim to its source before believing it. Legitimate claims lead to a PubMed entry, a journal’s Notice of Concern, the public record of the failed fosgonimeton trial, or the FDA’s compounding page, and they hold up once you get there. A certificate of analysis a seller wrote for itself is not independent verification, it is a document the seller chose to provide, so purity claims and “proven brain enhancer” language should be treated as marketing until a primary source backs them up.
What does “for research use only” mean in practice? It means the product is being sold legally as a laboratory chemical rather than as a medicine intended for people. That label is the entire legal foundation these products rest on, since sold for human use they would count as unapproved new drugs. The human use most buyers actually intend remains unapproved regardless of the label, which functions to sidestep medical regulation rather than to demonstrate the compound is safe or effective.
If supervision cannot make these peptides work better, why do FormBlends and HealthRX.com score higher than the research-chemical retailers? Because this checklist scores verifiability and trust, not effectiveness. No amount of clinical oversight turns a small foreign trial into robust proof, and no clinician can un-flag a paper a journal has already questioned. What the supervised tier adds is a licensed clinician and a licensed 503A pharmacy operating inside a documented transaction, candid disclosure of how limited the evidence actually is, and a relationship that continues past checkout, none of which the research-chemical tier offers or claims to.
Methodology and references
Providers were scored against eight criteria, weighted in this order: clinician oversight; licensed-pharmacy dispensing under 503A; verifiable identity and purity; candor about a small, mostly foreign, partly contested evidence base; operation inside a recognized legal framework; follow-up beyond the point of sale; pricing presented as a fair range rather than a lure; and whether claims survive a click to a primary source. The honesty criterion carried extra weight, because in a category this uncertain, a provider’s willingness to disclose the limits of the data is itself a meaningful safety signal. Price, shipping speed, and catalog breadth were deliberately excluded. Providers fall into two tiers that are not competing on the same axis, supervised medical telehealth and research-chemical retail, with the second tier ordered by general visibility rather than by any quality judgment, since relative purity cannot be independently confirmed from outside.
- Dolotov OV, et al. Semax regulates BDNF and trkB expression in the rat hippocampus. Brain Research, 2006. Animal study; a single intranasal dose produced “a maximal 1.4-fold increase of BDNF protein levels.” https://pubmed.ncbi.nlm.nih.gov/16996037/
- Gusev EI, et al. The efficacy of semax in patients at different stages of ischemic stroke. 2018. Human study of 110 stroke patients; plasma BDNF “remained high during the whole study period.” Non-blinded, Russian-language. https://pubmed.ncbi.nlm.nih.gov/29798983/
- Zozulia AA, et al. Efficacy and mechanisms of the peptide anxiolytic selank in GAD and neurasthenia. 2008. Controlled study of 62 patients; “the anxiolytic effects of both drugs were similar but selank had also antiasthenic and psychostimulant effects.”
- Filatova E, et al. GABA, Selank, and Olanzapine Affect GABAergic Neurotransmission Genes in IMR-32 Cells. Frontiers in Pharmacology, 2017. In-vitro; “Selank has no direct effect on the mRNA levels of the GABAergic system genes” alone.
- Sun X, et al. AngIV-Analog Dihexa Rescues Cognitive Impairment in the APP/PS1 Mouse via PI3K/AKT. Brain Sciences, 2021. Independent animal study; dihexa “restored spatial learning and cognitive functions.” No retraction or concern on this record.
- Notice of Concern regarding McCoy AT, et al. (2013 dihexa rodent paper), J Pharmacol Exp Ther. Notice of Concern issued 2021. A related 2014 HGF/c-Met paper from the same group has been retracted and is deliberately not used as evidence here.
- Fosgonimeton (ATH-1017) record, ALZFORUM. The prodrug built on the same mechanism dihexa targets failed its Phase 2/3 LIFT-AD Alzheimer’s trial, reported September 2024. Dihexa itself has no completed published human efficacy trial.
- U.S. FDA, Human Drug Compounding.
Are nootropic peptides actually safe to use?
Safety depends heavily on sourcing, dosing, and whether a real clinician is part of the picture. Most cognitive peptides in this space lack large-scale human safety trials, so a blanket “yes, safe” is not something anyone can honestly offer. What the evidence does show clearly is that contaminated or misdosed products, which show up often from unregulated online sellers, carry real risk. Working through a licensed, physician-supervised pharmacy remains the most defensible way to reduce that risk.
Do these peptides actually work, or is it mostly hype?
It depends on which peptide and what “work” is being asked to mean. Semax and selank have a reasonable body of human research behind them, largely from Russian clinical programs, pointing to modest effects on focus and anxiety. Others rest mainly on animal data or anecdote. Effects across the board tend to be subtle rather than dramatic, and individual response varies considerably. Any claim of transformative results is running ahead of what the evidence supports.
What are the most studied peptides for cognitive function right now?
Semax, selank, and dihexa draw the most attention in the clinical and preclinical literature on cognition. Semax has the longest human track record, originally developed in Russia for stroke recovery. Cerebrolysin has a genuine clinical trial base as well, though it is typically administered intravenously in a medical setting. BPC-157 is popular in biohacking circles, but its cognitive evidence remains almost entirely animal-based at this point.
Given all the crackdowns, where should someone actually buy these?
After the 2026 regulatory tightening, the gray-market research-chemical route became riskier, not safer. A physician-supervised compounding pharmacy, and FormBlends is one operating in this space, offers verified potency, sterility testing, and a licensed clinician reviewing the individual case. That accountability is the real practical difference now. Ordering from an overseas research-chemical vendor still happens, of course, but it amounts to guessing at purity and dose with no recourse if something goes wrong.
Written by Felix Zamora, wellness reporter. Last reviewed February 2026.
For readers’ general information. Medical decisions belong with you and a licensed professional.
Ventilation and Airflow Basics for a Sauna Heater: A Beginner Comparison