Energy, Cognition & Mood
Best Peptides for Memory & Cognitive Enhancement: Evidence (2026)
A clinical, evidence-graded ranking of the peptides marketed for memory — Cerebrolysin, Semax, Noopept, Dihexa and P021 — separating the disputed human RCT signal from preclinical mechanism, a retracted-data scandal, and pure marketing.
CerebrolysinSemaxNoopeptnootropic peptidesevidence-graded
The quick verdict
An honest, evidence-graded ranking of the peptides marketed for memory — and why no peptide is a proven way to make a healthy brain sharper.
- Best overall
- Cerebrolysin — The only peptide here with a large, genuine human-RCT base and a real cognition signal in vascular dementia and Alzheimer's — though efficacy is contested, sponsor-correlated, and confined to injured brains, never healthy enhancement (Grade A).
- Best value
- Semax — The best-evidenced synthetic peptide for the cost, with small human attention and short-term-memory data; the enhancement-marketed Dihexa and P021 have no human efficacy data at all, so they justify no premium.
- Best for Post-stroke or post-traumatic cognitive rehabilitation (clinician-supervised, not self-use)
- Semax — Its strongest human data are in acute ischemic stroke rehabilitation, where it raised plasma BDNF and accelerated recovery in a 110-patient trial — a genuine but single-country, under-replicated signal.
How we evaluated
We ranked each peptide strictly by the strength and volume of human evidence for memory or cognition as a measured endpoint — not fame, mechanism elegance, or evidence for unrelated conditions. Human trials outrank animal data, which outrank in-vitro and mechanism-only reasoning; anecdote and vendor copy do not count. Because none of these was tested for memory enhancement in healthy adults, we grade each by its best human data touching cognition and flag where that signal is contested, single-country, retracted, or absent. Evidence grades follow PeptideVox's A-to-D ramp (A = human RCT/meta; B = lower-tier human; C = preclinical only; D = anecdotal/marketing).
- Human memory/cognition evidence. Whether any published human study used a memory, attention or cognitive endpoint, and in what population — an injured/demented brain versus a well adult.
- Evidence quality & independence. Trial size, blinding, controls, region/language diversity, replication, and whether positive results correlate with manufacturer sponsorship or a data-integrity problem.
- Mechanistic coherence. Whether a plausible mechanism (BDNF/NGF/TrkB, synaptogenesis, anti-tau/anti-amyloid, monoamine/glutamate modulation) supports the claim, ideally with a human correlate.
- Safety & fit. Tolerability, route, serious-adverse-event signals, oncogenic or product-integrity risks, and uncharacterized long-term data.
- Regulatory honesty. FDA approval/compounding status and WADA exposure, stated plainly rather than glossed over.
Rating scale: 1-5 stars reflecting human-evidence strength for memory/cognition specifically: 5 = robust, independent human RCT data in the target population; 1 = mechanism or marketing only. Star ratings track the per-item evidence grade and its trustworthiness.
Last verified .
At a glance
| # | Name | Evidence | Rating | Best for | Pricing |
|---|---|---|---|---|---|
| 1 | Cerebrolysin | A | 2.5 | Understanding the ceiling of peptide evidence for cognition — a real, extensively studied drug for injured brains, not a memory aid for healthy ones | No US NDA; legal here only as a non-human research reagent; not a sourcing recommendation |
| 2 | Semax | B | 2.5 | Those wanting the best-evidenced synthetic peptide for cognition, with eyes open to its single-country, under-replicated evidence base | Research chemical / not FDA-approved; not a sourcing recommendation |
| 3 | Noopept (omberacetam) | B | 2.0 | People examining the honest ceiling of a much-marketed oral nootropic peptide — a real but thin clinical MCI signal, no healthy-adult proof | Research chemical / not FDA-approved; not a sourcing recommendation |
| 4 | Dihexa | C | 1.5 | Illustrating how a retracted-data scandal and a failed clinical prodrug demolish a heavily marketed enhancement claim — not for anyone seeking proof | Research chemical / not FDA-approved; not a sourcing recommendation |
| 5 | P021 | C | 1.0 | Understanding why an attractive preclinical mechanism, absent any human data, cannot support a memory-enhancement claim | Research chemical / not FDA-approved; not a sourcing recommendation |
Cerebrolysin
The only large human RCT base — but contested, and for injured brains
Cerebrolysin is the rare peptide product with a genuinely large human-RCT base, which is why it ranks first on evidence volume even though its verdict is best summarized as "studied extensively, proven inconclusively." It is not a single molecule but a porcine-brain-derived mixture of low-molecular-weight peptides (roughly 25%) and free amino acids (roughly 75%), produced by standardized enzymatic hydrolysis (code FPF-1070), and given parenterally only by slow IV infusion or intramuscular injection. Its most internally consistent positive cognition signal is in vascular dementia: a 2013 Cochrane review of six RCTs found benefit on cognition (MMSE and ADAS-cog gains) and a global-response advantage, and a 2024 network meta-analysis ranked it more effective than rivastigmine on ADAS-Cog. In mild-to-moderate Alzheimer's a meta-analysis of six placebo-controlled RCTs found a four-week cognitive benefit that was no longer significant at six months, with a sustained global benefit. Post-stroke and TBI series report benefit on composite endpoints, justifying Grade A for those indications. But three caveats gut its case. First, none of this tested healthy memory enhancement — every positive trial was in an injured or demented brain, so for healthy enhancement the grade is D. Second, the efficacy direction is sponsor-correlated: the manufacturer-independent 2023 Cochrane stroke review of seven RCTs found no benefit on death or dependence and a possible increase in non-fatal serious adverse events (RR 2.39), on the basis of which European stroke guidance advises against use. Third, it is parenteral and porcine-derived, carrying an anaphylaxis risk, and epilepsy is a contraindication.
Strengths
- The only peptide here with a large, genuine human-RCT base (dozens of trials, several meta-analyses)
- Most internally consistent cognition benefit in vascular dementia (2013 Cochrane, six RCTs)
- Modest four-week cognitive benefit in mild-to-moderate Alzheimer's and benefit on composite stroke/TBI endpoints
- Long clinical track record as an adjunct in injured/diseased brains (Grade A for those indications)
Weaknesses
- Grade D for healthy memory enhancement — never tested in a well brain
- Independent 2023 Cochrane stroke review found no benefit on death/dependence plus a possible serious-adverse-event signal (RR 2.39), and positive results correlate with manufacturer sponsorship
- Parenteral only (IV/IM), porcine-derived with anaphylaxis risk, epilepsy contraindication — not a casual nootropic
- Best for
- Understanding the ceiling of peptide evidence for cognition — a real, extensively studied drug for injured brains, not a memory aid for healthy ones
- Pricing
- No US NDA; legal here only as a non-human research reagent; not a sourcing recommendation
Source: Ziganshina et al., Cochrane CD007026 2023 (PMC10565895)
Semax
The best-evidenced synthetic peptide — small human attention & stroke data
Semax is a Russian-developed synthetic heptapeptide (Met-Glu-His-Phe-Pro-Gly-Pro), an ACTH(4-10) analog engineered to retain neurotropic activity while shedding the hormonal effects of the parent fragment. It is a registered prescription drug in Russia since 1994 and in Ukraine, but is not FDA- or EMA-approved. It ranks second here because, among the synthetic peptides, its human data come closest to a memory or attention endpoint. Its strongest clinical data are in acute ischemic stroke recovery: a 110-patient rehabilitation trial using 6,000 micrograms per day intranasally in two ten-day courses raised plasma BDNF and accelerated Barthel-Index and motor recovery, with a positive BDNF-to-Barthel correlation, and a Russian-literature meta-analysis of eight studies (n=654) reported neurological benefit. On cognition specifically, Kaplan and colleagues reported that intranasal Semax improved attention and short-term memory in healthy volunteers across a fatiguing workday, with EEG changes resembling nootropic drugs. The animal mechanism — robust, reproducible BDNF/NGF/TrkB induction plus monoamine modulation and penumbral neuroprotection in ischemia — is more consistent than the human cognition data. The critical caveat is source quality: the human signal is real but small, predominantly Russian-language, variably blinded, generated within a few laboratories, and never replicated to Western Phase 3 standards, and the independent Alzheimer's Drug Discovery Foundation review judges the cognition data limited, methodologically weak, and insufficient to recommend Semax for enhancement. That ceiling is why it earns B, not A.
Strengths
- Genuine human clinical pedigree in acute ischemic stroke recovery (raised plasma BDNF, accelerated Barthel/motor recovery)
- Small human studies actually measured attention and short-term memory, with the largest effect in fatigued subjects
- Robust, reproducible animal mechanism via BDNF/NGF/TrkB and stimulated monoamine tone
- Non-invasive intranasal route with CNS effects outlasting the minutes-long plasma half-life
Weaknesses
- Human data are small, predominantly Russian, variably blinded, and never replicated in large Western Phase 3 trials — capped at Grade B
- ADDF judged the cognition-specific evidence insufficient to recommend Semax for enhancement; no memory RCT in healthy adults
- Route-specific nasal irritation and a reported minor glucose signal in diabetics; research-chemical purity risk
- Best for
- Those wanting the best-evidenced synthetic peptide for cognition, with eyes open to its single-country, under-replicated evidence base
- Pricing
- Research chemical / not FDA-approved; not a sourcing recommendation
Source: Gusev et al., Zh Nevrol Psikhiatr 2018 (PMID 29798983)
Noopept (omberacetam)
One 56-day human MCI trial — healthy-adult use unproven
Noopept (omberacetam) is a synthetic proline-containing dipeptide ethyl ester (ethyl 1-(phenylacetyl)-L-prolylglycinate), a prodrug of the endogenous neuropeptide cycloprolylglycine, designed in Russia and grouped pharmacologically with the racetams; it is reported active at doses roughly a thousandfold lower than piracetam. Strictly speaking it is a peptide-derived small molecule rather than a long peptide, which is why it is orally active and crosses the blood-brain barrier. Its human evidence rests on a single cornerstone: Neznamov and Teleshova's 2009 study, a 56-day trial comparing Noopept 10 mg twice daily with piracetam 400 mg three times daily in patients with mild cognitive disorders of vascular or post-traumatic origin. Both arms improved on cognitive and mood scales, with the Noopept arm showing a favorable effect on anxiety, asthenic and affective symptoms. Supporting open-label experience is uncontrolled, and the dominant real-world use — cognitive enhancement in healthy adults — has no supporting controlled human trials, resting instead on extrapolation from rodent data. Preclinically it upregulates BDNF and NGF mRNA, engages HIF-1 cytoprotective signaling, modulates AMPA and cholinergic systems, and reduced amyloid-β-induced apoptosis with lowered tau Ser396 phosphorylation in cell models. It is graded B, not A, because the human evidence is small, single-country, active-comparator rather than placebo-controlled, unreplicated, without meta-analysis, and only 56 days long. A serious product-integrity problem compounds the risk: independent testing found U.S. supplements containing omberacetam at up to four times the pharmaceutical dose and frequently mislabeled or co-adulterated with other unapproved drugs.
Strengths
- A genuine 56-day human comparative trial in vascular/post-traumatic mild cognitive impairment, with cognitive and mood improvement
- Orally active and blood-brain-barrier permeable — a practical route unlike parenteral or intranasal peers
- Coherent preclinical mechanism (BDNF/NGF upregulation, AMPA/cholinergic modulation, anti-amyloid signaling in cell models)
- Favorable effect on anxiety and asthenic symptoms alongside cognition in the pivotal trial
Weaknesses
- Healthy-adult cognitive enhancement — its dominant real-world use — has no controlled human trials at all
- Human data are small, single-country, active-comparator (not placebo), unreplicated and only 56 days long — capped at Grade B
- Serious product-integrity risk: U.S. supplements found at up to 4x the pharmaceutical dose and adulterated; insomnia, irritability and increased blood pressure reported
- Best for
- People examining the honest ceiling of a much-marketed oral nootropic peptide — a real but thin clinical MCI signal, no healthy-adult proof
- Pricing
- Research chemical / not FDA-approved; not a sourcing recommendation
Source: Neznamov & Teleshova, Neurosci Behav Physiol 2009 (PMID 19234797)
Dihexa
Preclinical only — foundational papers retracted, prodrug failed in humans
Dihexa (PNB-0408) is a synthetic, blood-brain-barrier-permeable angiotensin-IV-derived oligopeptide developed at Washington State University, proposed to enhance cognition by potentiating hepatocyte growth factor (HGF) at the c-Met receptor and driving hippocampal synaptogenesis. It is among the most heavily marketed peptides for memory enhancement, which makes its evidence collapse the central honesty point of this ranking. No human trials of dihexa itself exist; its best effects — reversal of memory deficits in scopolamine-amnesia and aged rats, and synaptogenesis in vitro — are rodent and cell-system only. Worse, the central mechanism paper (Benoist et al. 2014, the source of the c-Met-phosphorylation and synaptogenesis data and the cited 65-picomolar HGF affinity) was formally retracted in April 2025 after a Washington State University investigation found falsified and fabricated data. The famous claims — "seven orders of magnitude more potent than BDNF," picomolar synaptogenesis — trace to that retracted source or to unverified extrapolations and cannot be relied upon. Human evidence exists only for a related molecule and it is negative: dihexa was deemed undruggable, so the optimized prodrug fosgonimeton (ATH-1017) was developed; it was safe in a Phase 1 trial of 88 subjects but failed its primary Alzheimer's endpoint in the Phase 2/3 LIFT-AD trial of about 315 patients (Global Statistical Test -0.08, P=0.70). Layered on top is a credible, unstudied oncogenic concern: c-Met is a well-characterized oncogene, and chronic potentiation of HGF/c-Met is a plausible cancer-promotion risk repeatedly flagged by the ADDF. Human safety of dihexa itself is essentially unknown, and dose/safety-margin extrapolations built on the retracted data are unreliable.
Strengths
- Blood-brain-barrier-permeable, orally/transdermally active in animal models — a favorable delivery profile in principle
- Reversed memory deficits in scopolamine-amnesia and aged-rat models (rodent-level evidence)
- Coherent proposed mechanism (HGF/c-Met potentiation, hippocampal synaptogenesis) that motivated a real clinical development program
Weaknesses
- No human trials of dihexa itself; foundational mechanism paper retracted in April 2025 for falsified data
- The optimized clinical prodrug fosgonimeton failed its Phase 2/3 Alzheimer's endpoint — negative human evidence for the class
- Credible, unstudied oncogenic risk via chronic c-Met potentiation; human safety of dihexa essentially unknown and not legally compoundable
- Best for
- Illustrating how a retracted-data scandal and a failed clinical prodrug demolish a heavily marketed enhancement claim — not for anyone seeking proof
- Pricing
- Research chemical / not FDA-approved; not a sourcing recommendation
P021
Preclinical only — zero human trials of any phase
P021 is a small synthetic ciliary-neurotrophic-factor (CNTF) mimetic (Ac-DGGLAG-NH2), derived by epitope-mapping CNTF's neurotrophic core and adamantylating it for stability and blood-brain-barrier penetration; it is orally bioavailable. It ranks last because its entire efficacy record is preclinical and largely confined to the single originating Iqbal laboratory. In rodent models of aging, Down syndrome and Alzheimer's, P021 raised BDNF, rescued dentate-gyrus neurogenesis, lowered GSK-3β-driven tau hyperphosphorylation, partially reduced amyloid-β, and improved maze learning — a mechanistically attractive package, but one that has never left animals. There are zero registered or published human trials of any phase. A useful check against overgeneralization: in a Cdkl5-knockout model, chronic P021 failed to raise BDNF or correct neuroanatomical defects, showing the effect is not universal even in rodents. On dosing, only animal figures exist (mouse dietary and aged-rat gavage ranges); no human dose, route or schedule has ever been established. On safety, investigators reported no weight loss, tumors or overt toxicity in roughly 18 months of dietary rodent dosing, and — unlike native CNTF — no anti-CNTF antibodies; but single-lab rodent absence of toxicity is not human safety, and theoretical immunogenicity and proliferative concerns remain unresolved. Marketing that presents P021 as a proven pro-memory or anti-Alzheimer's therapy is far ahead of an evidence base that contains no human data whatsoever, which is why it sits at the bottom of this ranking despite an appealing preclinical story.
Strengths
- Orally bioavailable and blood-brain-barrier-permeable in animal models — a practical delivery profile in principle
- Broad, coherent preclinical package: raised BDNF, rescued neurogenesis, lowered tau hyperphosphorylation, improved maze learning across several rodent models
- Long-term rodent dosing reported no tumors, weight loss or anti-CNTF antibodies (single-lab, animal-level reassurance)
Weaknesses
- Zero registered or published human trials of any phase — the weakest human evidence base in this ranking
- Efficacy record is largely single-lab, with at least one negative rodent model (Cdkl5-knockout) where it failed to raise BDNF
- No established human dose, route or schedule; human safety entirely unknown, with unresolved immunogenicity/proliferation concerns
- Best for
- Understanding why an attractive preclinical mechanism, absent any human data, cannot support a memory-enhancement claim
- Pricing
- Research chemical / not FDA-approved; not a sourcing recommendation
Frequently asked
Is any peptide proven to improve memory in healthy people?
No. Every positive human signal in this category comes from stroke, traumatic brain injury, vascular dementia, Alzheimer's or post-traumatic mild cognitive impairment — not from cognitively healthy adults. The Alzheimer's Drug Discovery Foundation judges the human evidence insufficient to recommend Semax for cognitive enhancement, and Noopept's healthy-adult use has no controlled trials at all. The most-marketed compounds for enhancement, Dihexa and P021, have zero human efficacy data of any kind. From a functional, root-cause standpoint the validated levers for memory — sleep, aerobic exercise, metabolic and cardiovascular health, omega-3 status, hearing correction and cognitive engagement — remain far better evidenced than any peptide here.
Which of these peptides has the most human evidence?
Cerebrolysin, by a wide margin — dozens of randomized controlled trials and multiple meta-analyses, which is why it ranks first on evidence volume and earns Grade A. But the direction of effect is contested. The most internally consistent positive cognition signal is in vascular dementia and mild-to-moderate Alzheimer's, where Cochrane and later meta-analyses found benefit. In acute ischemic stroke, however, the independent 2023 Cochrane review was null-to-negative and flagged a possible increase in non-fatal serious adverse events, and it noted that positive results correlate with manufacturer sponsorship. So Cerebrolysin is best understood as extensively studied but inconclusively proven, and it was never tested for healthy memory enhancement.
Why are Semax and Noopept graded B and not A?
Because their human trials are small, predominantly Russian-language, often non-blinded or active-comparator rather than placebo-controlled, generated within a few laboratories, and never replicated in large independent Western Phase 3 trials. Semax's clinical pedigree is mostly in acute stroke rehabilitation; Noopept's cornerstone is a single 56-day study comparing it to piracetam in vascular or post-traumatic mild cognitive impairment. Both showed benefit, but the source-quality ceiling — regional concentration, weak blinding, no meta-analysis, short duration — is exactly why they earn a B rather than an A. Russian regulatory approval is not equivalent to Western-grade proof.
Is Dihexa really 'more powerful than BDNF'?
That claim cannot be relied on. It traces to the Benoist et al. 2014 paper, which was formally retracted in April 2025 after a Washington State University investigation found falsified and fabricated data. The 'seven orders of magnitude more potent than BDNF' and picomolar-synaptogenesis figures derive from that retracted, fraud-tainted source. Dihexa itself has never been tested in humans. Its optimized clinical prodrug, fosgonimeton, was safe in Phase 1 but failed its primary Alzheimer's endpoint in the Phase 2/3 LIFT-AD trial. Layered on top is a credible, unstudied oncogenic concern: c-Met, the receptor Dihexa is claimed to potentiate, is a well-characterized oncogene.
What about P021 — is it a proven anti-Alzheimer's or memory peptide?
No. P021 is a small synthetic CNTF mimetic whose entire efficacy record is preclinical and largely from the single originating laboratory. In rodent models of aging, Down syndrome and Alzheimer's it raised BDNF, rescued neurogenesis, lowered tau hyperphosphorylation and improved maze learning — but there are zero registered or published human trials of any phase. There is also at least one negative animal datapoint: in a Cdkl5-knockout model P021 failed to raise BDNF or correct neuroanatomical defects, a useful check against overgeneralization. No human dose, route or schedule has ever been established, and its safety in humans is entirely unknown.
Are these peptides legal to use in the United States in 2026?
None is FDA-approved as a cognitive enhancer. Semax was removed from FDA 503A Category 2 and is scheduled for a Pharmacy Compounding Advisory Committee review on July 24, 2026 under docket FDA-2025-N-6895 — a pending, non-binding step, not an approval, and removal from Category 2 is not authorization to compound. Dihexa is not legally compoundable and faces its own separate PCAC review. Noopept is treated as an unapproved new drug subject to import alerts and is not a lawful supplement ingredient. Cerebrolysin has no US NDA and P021 is investigational only. All are sold as research chemicals of unverified purity, and for tested athletes all fall under the WADA S0 non-approved-substances catch-all, prohibited at all times.