Verified July 2026 · Cited to primary sources
The nootropic peptides marketed for focus, memory and mental resilience, graded by human evidence and legal status. The short answer: the two with real human data, Semax and Selank, are Russian-approved drugs, not FDA-approved, and their trials have never been replicated to Western standards.
Evidence for this goal
Semax and Selank are the only cognitive peptides here with genuine human data, and both are Grade C: approved as drugs in Russia, backed by real Russian clinical trials, but never replicated in a Western RCT. Neither is FDA-approved and neither has a legal supervised US route today. Treat them as early, foreign-evidence nootropics: promising in Russian studies, unproven to FDA standards, and gray-market to access.
How to read this comparison
Compounds are listed alphabetically, not ranked for this goal. The overall grade reflects the evidence described in each record. Evidence for one condition, population or product form does not establish benefit for every use. Affiliate availability never changes a grade.
A historical FDA survey of compounded drugs found 31% failed standard potency testing. What is actually in the vial is a separate question from whether the compound works, and a rubric like this exists to keep the two apart. See the full A to F methodology →
Evidence scope: Cerebrolysin sits at C: it has genuine human randomized controlled trials, including positive ones, but the overall body of evidence is mixed and the strongest signals are in the softer indications. In acute ischemic stroke, meta-analyses of randomized trials (over 1,700 patients pooled) failed to show significant superiority over placebo on core functional outcomes like the modified Rankin Scale and Barthel Index, though safety was neutral. The more convincing data is in vascular dementia: a randomized, double-blind, placebo-controlled multicenter trial in 242 patients showed a clear cognitive benefit (ADAS-cog+ improved 10.6 vs 4.4 points at 24 weeks, p<0.0001), and a meta-analysis pooled a beneficial cognitive effect. Real human RCTs, mixed results, foreign approval but no FDA approval, and methodological quality that varies by study. That is a textbook C.
Sources: Guekht et al. 2011, J Stroke Cerebrovasc Dis - Cerebrolysin in vascular dementia: randomized, double-blind, placebo-controlled multicenter trial (242 patients; ADAS-cog+ improved 10.6 vs 4.4 points at 24 weeks, p<0.0001) · Zhang et al. 2017, Biomed Res Int - Efficacy and Safety of Cerebrolysin for Acute Ischemic Stroke: Meta-Analysis of RCTs (7 trials, 1,779 patients; NO significant superiority over placebo on mRS/Barthel Index, neutral safety)
Cerebrolysin is not a single peptide but a standardized mixture of low-molecular-weight neuropeptides and amino acids produced by enzymatic breakdown of purified pig brain protein. It is proposed to act like a neurotrophic factor, supporting neuron survival, plasticity, and repair. It is given intravenously, usually in treatment cycles. Because it is a complex biological mixture rather than a defined molecule, its exact active components are not fully characterized.
Cerebrolysin is the most legitimately studied peptide in this batch and is a genuinely approved drug in several countries, which sets it apart from the research-chemical crowd. But the evidence is honestly mixed: the big stroke meta-analyses were neutral, and the stronger cognitive results in vascular dementia come from trials of variable quality. It is not FDA approved, so in the US it is research-only. Real drug abroad, real trials, real caveats.
See the evidence →Evidence scope: Dihexa is animal-only, and the picture is actually worse than typical preclinical uncertainty. The foundational rat studies showed dihexa reversing scopolamine-induced learning deficits and improving spatial memory in aged rats via HGF/c-Met, but the key 2014 mechanism paper has been retracted and the 2013 evaluation paper carries a formal expression of concern from the journal. On top of that, the one time this target reached humans at scale, it lost: fosgonimeton (ATH-1017), a brain-penetrant HGF/MET activator developed from the same dihexa lineage by Athira Pharma, failed its Phase 2/3 LIFT-AD Alzheimer's trial in September 2024, missing its primary and key secondary endpoints. So dihexa stays grade D on animal-only evidence, with the added honesty that its foundational data is integrity-flagged and its closest clinical relative failed in people.
Sources: McCoy et al. 2013, J Pharmacol Exp Ther - Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents (rat study; dihexa reversed scopolamine learning deficits). NOTE: this paper carries a 2021 journal Expression of Concern - read it with that caveat · ALZFORUM - Fosgonimeton (ATH-1017) therapeutics entry: documents the dihexa-derived HGF/MET drug and its FAILED Phase 2/3 LIFT-AD Alzheimer's trial (Sept 2024, missed primary and key secondary endpoints)
Dihexa was engineered from angiotensin IV to be orally active and cross the blood-brain barrier. It is reported to bind hepatocyte growth factor (HGF) and enhance signaling at its receptor c-Met, which drives dendritic spine growth (spinogenesis) and new synapse formation in the hippocampus. In cell models it was described as far more potent than BDNF at promoting neurotrophic activity. This entire mechanism story rests on preclinical work, some of which now carries formal integrity flags (see below).
Dihexa has one of the flashiest preclinical stories in the nootropic peptide world (orders of magnitude more potent than BDNF in a dish) and almost nothing solid to stand on. Its foundational rat papers carry a retraction and an expression of concern, it has zero human data of its own, and the closest clinical drug on the same HGF/MET target failed its Alzheimer's trial in 2024. Fascinating target, unproven and integrity-flagged evidence. Not something to treat as established.
See the evidence →Evidence scope: Humanin has a deep and consistent preclinical record but essentially no human efficacy data, which puts it at D. Its potent humanin analog S14G-HN prevented amyloid-beta-induced memory impairment in mice, [Gly14]-humanin protected spatial learning and memory in rats, and cell studies show it protecting neurons and synapses from Alzheimer-related insults. Humanin is also lower in the cerebrospinal fluid of Alzheimer's patients, which is suggestive biomarker evidence. But no completed human neuroprotection trial has been published, and the efficacy signal comes almost entirely from transgenic mouse and cell models, which have historically translated poorly to human dementia results. Strong biology, animal-only proof, so grade D.
Sources: Tajima et al. 2005, J Neurosci Res - A humanin derivative, S14G-HN, prevents amyloid-beta-induced memory impairment in mice (in vivo mouse study; S14G-HN prevented Abeta-induced spatial working memory impairment) · Zarate et al. 2019, Front Aging Neurosci - Humanin, a Mitochondrial-Derived Peptide Released by Astrocytes, Prevents Synapse Loss in Hippocampal Neurons (preclinical cell/animal study; humanin prevented glutamate-induced dendritic atrophy and synapse loss)
Humanin is a 24-amino-acid peptide encoded within the mitochondrial 16S rRNA (MT-RNR2) gene. It functions as a cytoprotective and neuroprotective signaling molecule, suppressing apoptosis and protecting cells against a range of stresses. In neurons it near-completely blocks cell death induced by Alzheimer's-related proteins and amyloid-beta, and it signals through pathways including STAT3, ERK1/2, and AKT. Circulating humanin declines with age, which has fueled interest in it as a longevity and stress-resistance peptide.
Humanin is one of the more scientifically interesting peptides here: a genuine mitochondrial-derived signaling molecule with a strong, consistent neuroprotective story in cells and animals, plus suggestive human biomarker data linking low levels to Alzheimer's. But interesting biology is not proven benefit. No human neuroprotection trial has been completed, the efficacy rests on mouse models that often fail to translate, and there is no human safety data for injecting it. Promising research target, not a proven treatment.
See the evidence →Evidence scope: Grade C: approved as a drug in Russia and backed by several small Russian human anxiety trials, including one comparing Selank to the benzodiazepine medazepam in 62 patients with generalized anxiety disorder (decision-tree rule 4 fired: foreign approval with early/weak human RCTs). No FDA-recognized Western RCTs have replicated these results.
Sources: Zozulia et al., Zh Nevrol Psikhiatr Im S S Korsakova 2008: Selank vs. medazepam in generalized anxiety disorder & neurasthenia (62 patients) · Medvedev et al., Zh Nevrol Psikhiatr 2014: anxiolytic effect and tolerability of Selank vs. phenazepam in anxiety disorders · Medvedev et al., Zh Nevrol Psikhiatr 2015: optimization of anxiety-disorder treatment with Selank · FDA 503A bulks list (Selank removed from Category 2; not on the July 2026 docket, and not on the agenda FDA published for the next PCAC meeting)
A tuftsin analog that modulates GABA and serotonergic signaling, raises BDNF, and slows enkephalin degradation. An anxiolytic/nootropic peptide developed at the Russian Academy of Sciences.
Selank is not FDA-approved (it is approved in Russia) and there is no legal compounding route today; it was removed from 503A Category 2 but is not on the July 2026 PCAC docket. Human evidence is Grade C: several small Russian anxiety trials, including one where it matched a benzodiazepine, but no Western replication.
See the evidence →Evidence scope: Grade C: approved as a drug in Russia with Russian clinical studies (stroke, cognition), but no rigorous Western RCT replication (rule 4 fired: foreign approval with weak/unreplicated RCTs).
Sources: FDA PCAC July 23-24, 2026 meeting (docket, July 24) · Federal Register FDA-2025-N-6895 (April 2026 503A reclassification)
Modulates BDNF and dopaminergic/serotonergic systems. A neuroprotective, cognition-enhancing nootropic peptide.
Semax is not FDA-approved (it is approved in Russia) and remains under FDA review. On July 24, 2026 the PCAC panel voted 8 to 5 (1 abstention) to recommend adding it to the 503A list for cerebral ischemia, migraine and trigeminal neuralgia, but that is advisory and the FDA has not acted. Human evidence is Grade C: real Russian trials, but no Western replication. For the anxiety-focused sister peptide, see our Selank page and our cognitive-peptides hub.
See the evidence →Evidence scope: There is genuine human trial data, which is more than most peptides here can claim, but the results are mixed. The pivotal Phase 3 MMPOWER-3 trial in primary mitochondrial myopathy (218 patients) missed both primary endpoints: the 6-minute walk test difference was -3.2 m (p=0.69) and the fatigue score difference was -0.07 (p=0.37). Separately, the dry-AMD program produced a positive Phase 2 ellipsoid-zone signal that was strong enough to carry a regulatory-accepted endpoint into the ongoing Phase 3 ReNEW study. So: a failed flagship indication alongside a live, better-looking one. Grade C reflects human trials of genuinely mixed results, not a clean win. It is not yet approved for anything.
Sources: Karaa A, et al. Efficacy and Safety of Elamipretide in Primary Mitochondrial Myopathy: MMPOWER-3 Randomized Clinical Trial. Neurology, 2023;101(3):e238-e252. · ReNEW: Phase 3 Study of Subcutaneous Elamipretide in Dry Age-Related Macular Degeneration (ongoing). · Genotype-specific effects of elamipretide in primary mitochondrial myopathy: post hoc analysis of MMPOWER-3. Orphanet J Rare Dis, 2024.
Elamipretide is a small cell-permeable tetrapeptide (D-Arg-dimethylTyr-Lys-Phe-NH2) that concentrates in the inner mitochondrial membrane and binds cardiolipin. By stabilizing cardiolipin and the associated respiratory-chain supercomplexes, it is proposed to improve electron transport efficiency, reduce reactive oxygen species, and restore ATP production in mitochondria that are dysfunctional. Because mitochondrial dysfunction sits underneath many age-related and rare diseases, it has been tested across mitochondrial myopathy, heart failure, and dry age-related macular degeneration.
The most legitimate science on this list, and still not a green light. Elamipretide is a real drug candidate with a plausible mitochondrial mechanism and clean tolerability, but its flagship mitochondrial-myopathy Phase 3 missed its endpoints. The dry-AMD program is the one to watch. Until a Phase 3 reads out positive and it earns approval, treat grey-market SS-31 as an unproven, unapproved injectable and keep expectations low.
See the evidence →| Peptide | Overall grade | Legal & safety | Typical cost | Verdict | |
|---|---|---|---|---|---|
| Cerebrolysin A peptide mixture derived from pig brain that is actually approved and used for stroke and dementia in several countries, just not the US. It has a real stack of human trials, but the quality is mixed and the biggest stroke analyses came back neutral. | Grade C | See evidenceSee the evidence for Cerebrolysin | |||
| Dihexa A lab-built angiotensin IV analog designed to supercharge brain synapse formation, and in rats it looked staggeringly potent. But it has never been proven in a human, and the drug company version built on the same target failed its big Alzheimer's trial in 2024. | Grade D | See evidenceSee the evidence for Dihexa | |||
| Humanin A tiny peptide encoded inside your mitochondrial DNA that acts as a cellular bodyguard, protecting neurons from Alzheimer-related damage in the lab. The cytoprotection story is strong in animals and cells, but no human neuroprotection trial has ever been completed. | Grade D | See evidenceSee the evidence for Humanin | |||
| Selank A Russian-developed anti-anxiety and nootropic peptide with real but non-Western human trials. In Russian studies it matched a benzodiazepine for anxiety without the sedation or dependence. | Grade C | See evidenceSee the evidence for Selank | |||
| Semax A Russian-approved cognitive and neuroprotective peptide with real but non-Western clinical data. | Grade C | See evidenceSee the evidence for Semax | |||
| SS-31 (Elamipretide) The most seriously studied peptide on this list, run through real Phase 3 trials. Its flagship mitochondrial-myopathy trial missed its primary endpoints, but its dry-AMD program is still advancing, so the honest verdict is mixed, not settled. | Grade C | See evidenceSee the evidence for SS-31 (Elamipretide) |
Reading this table: Grades refer to each compound's stated evidence scope, not every use in this category. Legal status and Safety are separate badges; Verdict is the compound summary. Affiliate availability never changes a grade. Full methodology.
FAQ
The two cognitive peptides with real human data are Semax and Selank, both Russian-approved drugs graded C on our scale. Semax is the cognition and neuroprotection peptide; Selank leans anti-anxiety and anti-fatigue. Both have genuine Russian clinical trials but no Western RCT replication, so they're promising but unproven to FDA standards, and neither has a legal supervised US route.
No. Both are approved as drugs in Russia, not by the FDA. Semax is on the July 24, 2026 PCAC docket; Selank was removed from the 503A Category 2 list but has not been scheduled for a PCAC meeting. Neither has a legal supervised US route today, so access is gray-market.
The human evidence is early and foreign. Semax has Russian trials for stroke recovery and cognition; Selank matched a benzodiazepine for anxiety in a Russian study. Both are Grade C: real early human data, not the multiple Western RCTs that would earn an A or B. We wouldn't call their cognitive benefits proven to FDA standards.
Read the evidence
Before you buy anything
Other goals
The monthly peptide evidence brief.
What the research and the FDA actually say, in one short email a month. Unsubscribe anytime.
No spam. We never sell your email. Editorial policy.