How the grading works.
PepCue exists because most peptide writing online is marketing wearing a lab coat. Our rule is simple: every claim points back to a real source, or it doesn't ship. This page is the whole model, including the parts that make us look worse.
Evidence over hype
Every profile leads with what has actually been published in humans, and says plainly when the answer is 'not much.' Preclinical-only compounds are labeled preclinical-only, no matter how good the mechanism story sounds.
Real citations, always
Every factual claim resolves to a real paper, FDA document or trial registry. Nothing is a plausible-looking guess. If a value isn't verified, we don't print it.
Regulatory reality, not a vibe
Alongside the evidence grade we show the one fact that isn't debatable: whether a compound is FDA-approved, in registered trials, or research-only.
No affiliate codes
We don't sell peptides, rank vendors for money, or take kickbacks. There is nothing to buy here, so the incentive to shade the evidence doesn't exist.
Never a dose
PepCue never tells anyone what to take or how much. Dosing is a conversation for you and a licensed provider. We explain; we don't prescribe.
The six axes every compound is scored on
Each compound is rated 0 to 5 on six independent axes. They are deliberately separate, because the most common error in this field is letting a strong score on one axis quietly stand in for another. A compound can have excellent mechanism confidence and zero human evidence at the same time, and collapsing those into a single impression is how a rat study becomes a marketing claim.
How much controlled human data exists. A 17,000-person randomized outcomes trial scores at the top. A handful of uncontrolled case reports scores near the bottom. No human data at all scores zero, regardless of how strong the animal work is. This axis carries the most weight in the final score.
How thorough the animal and cell work is. This can be high while human evidence is zero, which is exactly the pattern behind most research peptides, and precisely the gap the tier system is designed to make visible.
How well the biological pathway is actually characterized, as opposed to plausibly asserted. A receptor with a named target, a measured binding profile and reproducible downstream effects scores high. A proposed mechanism that independent labs could not reproduce scores low.
Not how safe a compound is, but how well its safety is known. An approved drug with post-market surveillance scores high. A compound with no human toxicology scores at the floor, because unknown is not the same as safe.
Whether a regulator has actually reviewed the compound. FDA approval scores at the top; registered trials score in the middle; research-use-only material scores at the bottom. Approval elsewhere but not in the US sits between.
Whether the studied outcome is the one people actually care about. A compound proven for a narrow orphan indication does not automatically transfer to the use it is marketed for online.
How the six axes become a letter
The axes are combined into a weighted score out of 100, then mapped to a letter. Human evidence carries the heaviest weight, followed by safety clarity and regulatory clarity. Mechanism confidence and preclinical depth carry real but smaller weight, because a mechanism is a reason to run a trial, not a substitute for one.
Two further inputs are derived rather than authored. Evidence consistency rewards findings that have been replicated. Claim risk is a penalty: when a compound is very widely discussed but has thin human evidence, the gap between what is said about it and what has been shown is itself a signal, and it pushes the score down. Popularity can only ever lower a grade on PepCue. It can never raise one.
A compound with no authored score is not given a letter. It appears on the board in a “review pending” group instead. Printing an unearned F would be exactly the kind of invented value this page promises not to print.
A worked example: BPC-157
BPC-157 is the most useful compound to walk through, because it is simultaneously one of the most popular research peptides and one of the least proven in humans. Here are its actual live scores, pulled from the same data that renders its profile:
Read across the row and the story is obvious: the preclinical literature is genuinely deep, the mechanism is reasonably characterized, and human evidence and safety clarity are close to the floor. That is not a knock on the compound. It is an accurate description of a molecule that has been studied intensively in rodents and barely at all in people, and it is the single most important thing a reader can know before deciding what to do with it. See the full profile at /p/bpc-157.
How claims get a verdict
Separately from the tier, we audit the specific claims that circulate about a compound in marketing copy and forums. Each gets one of six verdicts. The verdicts describe the state of the evidence behind a claim, never an invented study result:
The distinction between unverified and false matters more than any other on this site. Most claims about research peptides are unverified: nobody has run the trial. Calling that “false” would be as dishonest as calling it proven.
What we refuse to publish
A methodology is defined as much by its refusals as its rules. We do not publish doses, protocols, cycles or titration schedules for research compounds. Where an amount appears in a calculator it is arithmetic on numbers the reader entered, and where a schedule appears it is a reproduction of a manufacturer's FDA-approved label, attributed as such.
We do not name or rank vendors we have not verified, which is why the supplier page teaches evaluation criteria rather than publishing a leaderboard. We do not print a certificate of analysis as a safety guarantee, because a CoA is a chemistry document and says nothing about sterility or endotoxins.
And we do not invent citations. Every PMID on this site is checked against the live NCBI record by an automated audit that runs against the whole corpus; anything that fails to resolve does not ship.
Independence, and how to correct us
S–F tier boards and evidence-maturity scores aren't a novel idea. Several sites in this space use a similar shape. What we control is whether ours is honest: the full scoring model is on this page, not hidden behind a black box, and every input is disclosed per compound on its profile.
This reference and the PepCue app are built by the same team, and we're upfront about that link. The grades are governed by the scoring model alone: a grade is never adjusted to make the app, or any product, look better. If a compound's evidence is thin, the profile says so, full stop.
Grades are not permanent. They move when the evidence moves: a published trial, a new approval, a failed endpoint, a withdrawn drug. Content carries a reviewed date, and every page has a link to flag an error. If you can point to a source we got wrong, we would rather hear it than be wrong quietly.
Cite this page
PepCue. “Methodology: how PepCue grades the evidence.” PepCue, reviewed June 1, 2026. https://www.pepcue.app/methodology.
Start with the full index, sort by the evidence board, or browse by goal. When you're ready to track a protocol with your provider, that's what PepCue is for.