1. Start here
PeptidesDB is a reference index of published peptide research. Every number on it — the effect scores, the evidence grades, the study summaries — is either pulled from primary literature or derived from that literature by the Librarian, the automated pipeline that reads new papers each day. Nothing is a vote, and nothing is a reader submission.
That shapes how to read the whole site. Everything here is framed as “what the papers say”, which is a different claim from “what works” and a very different claim from “what you should take”.
Peptides indexed here are research compounds. Nothing on this site is a dosing instruction, a diagnosis, a treatment, or a recommendation for personal use — including the calculator, which does arithmetic on numbers you supply and takes no view on whether you should use them.
2. Find a peptide
The peptide library is where most visits start. It filters by evidence grade, effect category and research footprint, and each tile carries a summary and a miniature effect chart. Opening one gives you the full profile: what the compound is, what has been studied, the indexed studies themselves, regulatory status where it exists, and the adverse events reported to FDA FAERS.
Two shortcuts around it. Category hubs group compounds by what they are studied for rather than by name, and Compare puts two profiles side by side — useful when the question is which of two compounds has the literature, not whether either does. Protocols covers combinations that get reported together, with the constituent research attached to each.
3. Read an effect chart
Every peptide page carries a seven-axis chart scoring these categories from 0 to 5. The scores are assigned by the Librarian from the indexed studies, trials and label data for that compound. No human sets them and no vote moves them.
| Axis | Tracks |
|---|---|
| Healing & recovery | Tissue repair, tendon and wound work, post-injury recovery |
| Muscle & body comp | Lean mass, strength, body composition |
| Metabolic & appetite | Glucose handling, appetite, weight, lipids |
| Neuro & mood | Cognition, mood, neuroprotection |
| Sleep & circadian | Sleep architecture, circadian timing |
| Inflammation & immune | Inflammatory and immune signalling |
| Skin & aesthetics | Skin, hair, collagen and cosmetic endpoints |
Read the shape, not the individual points. A tall axis means that category is well represented in the research for that compound — not that the compound works, and not how well it works. A chart can move when new studies land, so a shape is a snapshot of the literature on the day you look at it. Follow it through to the studies before concluding anything from it.
4. Read an evidence grade
Every peptide carries a letter grade from A to D. A is human clinical evidence, B is early human evidence, C is preclinical work only, and D means nothing is indexed yet. The grade is computed from the study rows we hold — their evidence type, study type and registered trial phase — and recomputed as the Librarian ingests new work.
A grade describes the strength of the evidence base and nothing else. It is not a safety rating, not a recommendation, and not a measure of popularity: a compound can be graded A and still be entirely unsuitable for you, and a D means only that nobody has published, not that the compound is inert. The full definition of each tier is on the editorial policy page.
5. Follow the research
The research library is every indexed study in one feed, filterable by peptide, evidence type and source. Each row shows the title, the journal, what kind of study it is, and — where the Librarian could write one — a short summary carrying an AI analyzed label. The link to the original record is always on the row. Cite that, not us.
Next to each summary is a summary depth figure. It is the most misreadable number on the site: it describes how much of the paper reached us, not how good the paper is. A low figure usually means the feed handed us little more than a title, so our summary is thin and you should read the source before relying on it — plenty of excellent papers score at the bottom of that scale. The full explanation is on the editorial policy page.
6. News, signals and the week
Three surfaces sit outside the peer-reviewed corpus, and they are kept deliberately separate from it. News is coverage from established outlets. Signals is community discussion — anecdotal by construction, labelled as such, and never used to establish a fact or move a score. This week is the digest: what landed in the last seven days across all of them, in one timeline.
The blog is the one place on the site where we write at length rather than index. Posts carry a byline, a review date and a link to the author profile.
7. Do the arithmetic
The dose calculator converts between vial strength, reconstitution volume, concentration and syringe units. It is arithmetic and nothing more: it does not know what compound you typed in, does not check whether the number is sensible, and takes no view on whether you should use it. For the mechanics behind those numbers, the blog covers reconstitution, bacteriostatic water and injection technique in full.
Two directories complete the tools. Trackers follows GLP-1 supply and the regulatory position on compounded peptides, each statement dated and sourced. Vendors is a directory with certificates of analysis where we have them; inclusion is not an endorsement and no vendor pays to appear.
8. Check our work
Every action the Librarian takes is logged in the activity ledger — what it read, what it changed, how confident it was, and which sources it cited. The ledger is public because a machine that writes summaries is only trustworthy if you can audit it. Third-party testing sets out what an independent batch-testing programme here would publish, and is honest that none of it has started.
If a summary misrepresents its paper, or a page describes something that no longer exists on the site, tell us — corrections are read first and answered fastest. The editorial policy and AI disclosure sets out exactly which parts of this site are produced by a machine and where the line is drawn.